Engineering and testing of a structual joint for honeycomb panels in the VIP completion

Table of Content

Zusammenfassung

Abstract

Preface

Table of Content

List of Tables

Table of Figures

Nomenclature

List of Abbreviations

1 Introduction
1.1 BizJet International Sales & Support, Inc
1.2 Scope

2 Panel Joint techniques
2.1 Introduction
2.2 T Joint
2.2.1 Option 1
2.2.2 Option 2
2.2.3 Option 3
2.2.4 Use-Value Analysis T Joint
2.3 L Joint
2.3.1 Option 1
2.3.2 Option 2
2.3.3 Option 3
2.3.4 Option 4
2.3.5 Use-Value Analysis L Joint

3 Test arrangement
3.1 Introduction
3.2 Test equipment
3.3 Test requirements
3.4 Testing procedure

4 Test phase
4.1 Introduction
4.2 First test run
4.2.1 Test coupons
4.2.2 Evaluation of the test results
4.3 Second test run
4.3.1 Improvements
4.3.2 Evaluation of the test results
4.4 Third test run
4.4.1 Improvements
4.4.2 Manufacturing
4.4.3 Evaluation of the test results
4.5 Bending test
4.5.1 Introduction
4.5.2 Test results

5 Conclusion
5.1 Results
5.2 Outlook

List of References

Attachment A

Attachment B

Attachment C

Attachment D

Attachment E

Attachment F

Attachment G

Attachment H

Attachment I

Attachment J

Zusammenfassung

Name

Felix Pinkepank

Thema der Bachelorthesis

Engineering and testing of a structual joint for honeycomb panels in the VIP completion

Stichworte

Bizjet, Honeycomb Verbindung, Mortise und Tenon, Test, Wertanalyse, Panel Pins, Kabinenmonumente, Klebstoff, klebverbindung

Kurzzusammenfassung

Diese Arbeit befasst sich mit der Entwicklung einer neuen Verbindungstechnik für Honeycomb Panels. Dabei werden Techniken diskutiert und auf ihre Kompatibilität für Mortise und Tenon geprüft. Die ausgewählte Technik wird in verschiedenen Stufen der Entwicklung getestet und analysiert. Dabei werden Herstellungsprozesse optimiert um das Projektziel eine nicht nur stärkere, sondern auch effektivere Verbindung zu erreichen. Fuer dieses Projekt wird die aktuelle Verbindungsmethode von Panel pins als Vergleichsmethode benutzt. Testergebnisse werden vorgestellt und mit Hilfe von Diagrammen und Bildern analysiert.

Name

Felix Pinkepank

Title of the paper

Engineering and testing of a structual joint for honeycomb panels in the VIP completion

Keywords

Bizjet, honeycomb panel Joint, mortise and tenon, cabinet, panel pins, test plan, process specification, adhesive, adhesive joint

Abstract

This paper is a design analysis for the development of a new structural joint for honeycomb panels. Therefore different techniques will be discussed and checked for the compatibility with mortise and tenon. The selected technique will be tested and analysed in different stages of development. Additionally the manufacturing process will be optimized to achieve the goal not only to obtain a stronger but also a more effective joint. For this project the current joint with panel pins will be the method for comparison. The test results are presented and will be analysed with diagrams and figures.

Abstract

The method of pins as a joint for two honeycomb panels is the most common in the aeronautical business due to the high reliability and strength. Both honeycomb panels do not need to be cut to use pins as a joint. Therefore this method has its big advantages when a company has to manually cut and drill the honeycomb panel.

Since BizJet International invested in a full automatic CNC machine, this valuable resource raises the question whether panel pins are still the most effective way to build cabinets. Different ways for panel joints were discussed and a concept for a potential more effective way will be elaborated and tested. In preparation for the engineering of a new panel joint which would change the industry standard of BizJet International, design standards and guidelines for developing test coupons are analyzed and provided.

Following the first designs of test coupons in the CAD system Solidworks, is a company test to proof the strength for tension and shear loads. Furthermore a value analysis will be created to proof the economic benefit of the developed joint. The progress of this research and development project will be shown based on testing for each stage of development and detailed analysis of failure modes with diagrams and pictures with the goal to achieve a more effective way of joining honeycomb panels.

Preface

This thesis is the result of six month of work for Bizjet International in Tulsa, Oklahoma.

Since this project included the process from engineering and testing through to certification and process specification, I could get an extensive insight in the VIP business and the requirements for setting new standards in the aeronautical industry.

Foremost, I would like to express my sincere gratitude to my advisor Prof. Dr.-Ing. Gordon Konieczny for the continuous support. His guidance as an experienced professor helped me throughout my internship and the including bachelor thesis.

All in all it was a great experience to accomplish my internship at Bizjet International in the USA. I had a great time during my project and learned a lot about the practice as an engineer.

Therefore I would like to thank Curtis Brown for helping me with the unfamiliar CAD system Solidworks and Frank Merigliano who gave me a lot of helpful tips in engineering tasks. That integrating a German student into an American company is not always easy but possible has been proven by Shannon Carter and Brett Jackson who helped me organizing and structuring my internship abroad.

In my daily work I have been blessed with friendly and helpful colleagues that made this to an experience that I will always remember.

I would like to express my deepest gratitude to my industrial advisor B. Eng. Robert Deal for the trust, the insightful discussion and the providing of valuable advices throughout my time at BizJet. His experience and incredible knowledge of the aviation industry as an engineer and DER for the FAA made this project and its outcome possible.

Last but not the least; I would like to thank my family: my parents Anne and Ulrich Pinkepank and my three brothers who were always supporting me and encouraging me with their best wishes.

1 Introduction

1.1 BizJet International Sales & Support, Inc.

The VIP and Business sector is a large and still growing market in the aviation industry. For Lufthansa Technik, one of the biggest companies for maintenance, repair and overhaul tasks (MRO), BizJet is an affiliated company accountable for the completion and maintenance of VIP and Business Jet interiors.

Together with Hamburg, Bizjet, located in Oklahoma, designs and engineers complete cabin monuments for the ACJ A318, A319 and the Boeing BBJ series. These monuments, consisting cabinets, galleys and lavatories, are manufactured in-house from the CAD model through to the final assembly. Bizjet has grown very fast in the past years and nearly doubled the number of employees in four years. Not only the quantity of the company is increasing, but also the quality is increasing due to better machinery and an enhanced LEAN production. The main investment in new machinery was the high tech. CNC-machine. The CNC-machine enables the possibility to cut complex structures in the honeycomb panel as opposed to the old stationary saw which was time-consuming and imprecise.

This new production process could be a precursor of a more effective way to connect panels. Therefore, different panel joints and test methods were discussed to figure out the right way to proof the structure and fulfill the regulations of the FAA.

The following chapters will show the different panel joint methods. These are constituted of Bizjet standards and benchmarking. Thereafter the requirements for the test and the test arrangement will be presented. Subsequently the test phase will show different stages of development with analyses of the important samples. With the assistance of diagrams and pictures of the failed test coupons the results will be discussed and ideas for improvement will be presented.

1.2 Scope

The purpose of this bachelor thesis is to ascertain the effectiveness of mortise and tenon as a technique of manufacturing monuments for VIP Jet interiors.

Therefore the engineering and the applicable tests have to be designed to show compliance with current regulations.

The progressing steps of the research and development project shall be described and the test requirements and accomplishment will be elaborated. Each test will be analyzed separately with ideas and implementations for improvements.

2 Panel Joint techniques

2.1 Introduction

The first step in this project is to find an appropriate joint technique that has the potential to be better than panel pins relating to weight, strength, time exposure, manufacturability and costs. Whereas with panel pins there is no variability with the joint technique in T and L joints, there are different ways to joint two panels with the CNC cut method. Therefor different methods both for the T joint and the L joint were evaluated and discussed.

2.2 T Joint

The T joints are ordinarily the stronger joints because they are more resistant to shear loads and the conjunction offers better transmission of tension loads. For the second and third test only T joints will be tested in order to stay within a reasonable expense with at the same time proper results.

2.2.1 Option 1

The first option is designed to provide maximum strength with minimum amount of adhesive used. Figure 1 shows the panel joint with eight MaT. The dimensions of the MaT are chosen to provide good results in shear and tension tests however not weaken the perpendicular panel. Each mortise of course weakens the panel and makes it prone to a failure caused by bending. Each tenon has to be cored out manually and the skin has to be left intact to provide maximum results for strength. Therefor this method would have possible human errors in manufacturing. Additionally the time exposure of one panel would be massive if every tenon needs to be cored out manually.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1: Option 1 with coreless tenons

The longer the panel with tenons is, the more difficult it will be to align the panels. The missing core between the tenons will reduce the stiffness drastically especially when the skin is fibreglass- or carbon-fibre-reinforced polymer. Skin made of aluminium has enough stiffness to ensure proper alignment of the MaT but thus the majority of cabinets are made out of Norbond (“NB-220-155-xxx” with an e-glass face sheet) or Teklam (“NP2G1-02-xxx” with a phenolic face sheet), the problem of stiffness can be critical for the manufacturability. Another problem will be the injection of the mixed adhesive so that the skin bonds in the mortise safely. A major issue with that is that the minimum tool diameter on the CNC machine is 0.125 inches and the thickness of the skin is 0.02 inches which would lead to gaps between the mortise and the tenon on each side with a size of approximately 0.04 inches. This makes it very hard to inject the adhesive with pressure to ensure that every small gap is filled with adhesive. It also tends to an adhesive failure if the distance between the skins is too high. The details of the adhesive properties are shown in Chapter 4.2.1.2.

2.2.2 Option 2

The second option for a T Joint consists of long MaT with core inside the tenons. This Joint (visualized in Figure 2) would eliminate the issue of option 1 with instable tenons but would need way more adhesive in the mortise to bond the panels correctly. Every additional gram of adhesive adds up quickly in a cabinet with many MaT and can be a critical factor for designing future cabinets. The advantage of this joint would be the easy and fast manufacturability due to a less amount of MaT. Transverse shear will have the best results with this option, but the critical allowable for most panel joints is the tension allowable. Another factor is the long cuts into the panel which makes the panel vulnerable for bending. In this case another important test would be the long-beam bending test to show that the panel is still strong enough.

Abbildung in dieser Leseprobe nicht enthalten

Figure 2: Option 2 with long MaT

Flammability has to show that the large amount of adhesive does not burn. This has to meet FAA requirements 25.853, “Compartment interiors”.[1] These requirements are for flammability testing of materials used for interiors of the airworthiness standards: transport category airplanes.

2.2.3 Option 3

The third option shown in Figure 3 is a mix out of the first and the second option. It has double the length as the width and it is not cored out for the reason of simple manufacturing. Although this might be weaker in strength than option 2 and need more adhesive than in option 1 it got presumably still enough strength in longitudinal shear to meet current FAA regulations.

Abbildung in dieser Leseprobe nicht enthalten

Figure 3: option 3 with panel width related MaT

The tension strength which is the crucial factor will have the same or even higher strength than the other options. The number of MaT per panel can be evaluated easier cause they can be tested the same way as panel pins. Research showed that MaT in the wood technology got mostly about 1/3 the thickness of the length. The difference with honeycomb panels is, that the thickness is given by the size of the panel whereas the width of the wood tenons are about 1/2 as thick as the panel.

2.2.4 Use-Value Analysis T Joint

For the decision of the right way to go in the first tests the use-value analysis will rate each option in different categories. The emphasis for each category is numbered between 1 and 5 points. The rating for each option is rated between 1 and 4 points. To evaluate each option and compare it with the other the rating and the emphasis are multiplied and adds up to a final result.

In the aviation industry the weight always is an important factor for fuel consumption, maximum take-off weight and ultimate loads. Although the design is for the VIP completion, the requirements from the FAA are the same. Not only the weight of the aircraft is increasing, but also each cabinet has higher requirements for the design and its structure. For example 10 grams more on each MaT would lead to a total weight with approximately 100 MaT of 1 kilogram. The weight is closely linked to the material costs because it is mainly driven by the adhesive that is used for bonding the two panels together. The different adhesive materials that can be used are described in Chapter

4.2.1.2. The costs for the material got the least input in the use-value analysis because the increase of the costs for each cabinet will be minor. The adding or subtracting costs for exchanging the panel pins with adhesive shall be evaluated. The manufacturability describes how simple the steps of producing the joints are. This includes the demands on the human and the CNC. Whereas the CNC got the capability of doing complex shapes the steps done by humans is the key to an easy manufacturability. The best manufacturability would be to get everything done by the machine. The time exposure is measured in the time needed for completing a cabinet using MaT joints in comparison of panel pins. The last but very important factor is the strength of the joint. Although the strength has to be tested to get accurate values the advantages and disadvantages that would occur can be discussed and considered.

Abbildung in dieser Leseprobe nicht enthalten

Table 1: Use-Value Analysis for T Joints

The results of the use-value analysis can be inspected in Table 1. Option 1 with the coreless tenons got the best results in weight and material costs because the amount of adhesive is very low. Additionally the strength and therefor the allowables will be quite high. But the good balance between strength and weight would lower the easiness of manufacturing a cabinet. The steps to provide the right shape of the tenons would consume more time than doing a panel pin. Additionally the alignment would consume too much time due to the lost stiffness in the face sheets of the tenons and will need a patient manufacturer.

The second and the third option are pretty similar in the way they are manufactured. The only difference is the length of the panel and ensures a fast manufacturing. If the mortises got filled with pressure the manufacturer would not need to remove the pressurized adhesive gun as often as in option 3. This would decrease the amount of steps needed for one joint and therefor decrease the time exposure for all the cabinets. The weakness of the long MaT is in the strength. The susceptibility to a bending failure is too high and would therefor cost additional stress engineering for every cabinet to proof the resistance for a bending failure. The use-value analysis has resulted that option 3 has the best qualities for the panel joint and will be the element for the on-going researches.

2.3 L Joint

The L joints are normally the critical factor for laying out a cabinet. The allowables are significantly smaller than the allowables for the T joint. The techniques that can be used are similar to the T joints with one additional technique which is only possible on L joints. As compared with the T joints there are different requirements for option 1 and 3 concerning the bonding techniques. The required FAA regulations are described in Chapter 2.3.1 and Chapter 2.3.3. In recent projects of Bizjet International some of the L joints with panel pins were too weak. Especially with pins next to a floor fitting are vulnerable and need to be controlled extensively.

Abbildung in dieser Leseprobe nicht enthalten

Figure 4: Critical transverse shear on an L joint

Therefor an additional IRR[2] had to be provided to reinforce the L joint to withstand the ultimate loads in emergency landing conditions[3]. The critical value on this joint will be the transverse shear in the direction seen in Figure 4. The red arrow marks the pulling direction and it is visible that there is no honeycomb to absorb the shear forces in this directions. The adhesive will be the point of failure in this test and the allowables will probably be weak.

2.3.1 Option 1

The first option is the equivalent of the selected T joint. It has the same dimensions for the MaT with the mortise open on one side. Because the mortise is open to one side, the skin is even more affected with losing stiffness. Figure 5: option 1 with regular sized MaT shows the MaT joint with the open honeycomb on the mortise panel. When the panels are bonded together the adhesive will shrink and pull the mortise skin into the joint. The result would be an uneven surface which would have to be corrected with Aerok9111R. This adhesive is common for edge filling, edge protection and smoothing surfaces. This issue is only affecting the appearance of the panel and can be neglected in regard to structural requirements.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5: option 1 with regular sized MaT

Not only the losing stiffness is a problem but also this joint is not completely closed. This means that the adhesive is not only in the joint but also squeezes out on the side with the honeycomb. This could lead to an issue with the regulations for flammability. The appropriate part for this is the “FAR 25.853: Compartment interiors”[4]. The methods of compliance (MOC) with the flammability requirements are explained in this chapter respectively in Appendix F to part 25[5]. Since the main focus of this project is the shear and tensile strength of the MaT joint these regulations are secondary.

2.3.2 Option 2

The second option got some properties of the first option in Chapter 2.2.1. The core of the tenon will be removed and only one side of the tenon has a CNC cut shape. The other side ([1]) seen in Figure 6 has a continuous shape with the width of the panel with the mortise. The reason for that is to cover the exposed honeycomb of the panel which would lead to a very big bonding surface. Also the step of edge filling would not be needed in this option. The edge filling of the exposed honeycomb is to prevent moisture and physical damage to the honeycomb. The biggest issue will be the same as in Chapter 2.2.1 with the losing stiffness because of the removed core.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6: option 2 with coreless tenon

Although the problem is not only alignment in this option it is the manufacturing side that will cause difficulties. The losing stiffness of the side [1] has to be clamped to the other panel while the adhesive is curing. Therefor a clamping device would be needed to get a smooth surface. Since this process is nearly impossible especially without the possibility to use gravity as a force to press both panels together the big advantage in tensile strength becomes secondary. This technique might be a solution with panels that have aluminium skin so it does not lose the whole stiffness with a removed core. Unfortunately the majority of cabinets at BizJet consist of non-metallic skin and therefor this project is concerned with these types of panels.

2.3.3 Option 3

Similar to option 1 this possible L joint has the same shape with a longer tenon and thereby a longer mortise. The aspect of the shrinking adhesive in this joint can be critical. Whereas it was only a small lowered skin in option 1 it can be a depression which affects the strength of the joint. The cause for this is most likely the lost stiffness on the skin that is under the mortise. This would result in a time consuming correction for the surface on the panel. A positive aspect for this technique is that quality assurance is easier to ensure because this joint has a low quantity of MaT.

Abbildung in dieser Leseprobe nicht enthalten

Figure 7: option 3 with long MaT

With the same characteristic as the Option 2 in Chapter 2.2.2 it is faster to fill the MaT with adhesive due to the lower quantity. The pressurized gun will be triggered less and therefore the amount of cleaning he oozed adhesive is decreased.

2.3.4 Option 4

The last option for an L joint is an idea that came up with benchmarking and comparison to what already exists. That MaT are already used in different companies is not a secret and was one of the impulse for the beginning of this project. Since BizJet has an a318 and a 737-800 in the hangar it was possible to do a benchmarking on the components of these aircrafts.

The result of the benchmarking was that Boeing got MaT joints to connect the panels for the baggage bins. Figure 8 shows the technique that Boeing uses for the baggage bins and it is visible that they use a CNC machine for the panel designs. This CNC cut allows them to give the L joints a T joint shape.

Abbildung in dieser Leseprobe nicht enthalten

Figure 8: picture of Boeing 737-800 baggage bin

This design has a lot of advantages for example the additional strength due to the mortise pocket and the larger surface area for the adhesive. Even though it is not comparable with a T joint in strength the allowable will be approximately between the L and the T joint allowable. Additionally the tenon in the Boeing baggage bins got different length adjusted to the stress distribution inside the panel. The more load the bigger the MaT. Another advantage is the flammability certification in this joint because the adhesive is inside the joint and has no surface area on the outside. The concept that has been discussed shown in Figure 9 is slightly different from the MaT of the baggage bins. The mortise has somehow the same shape but the length is equal in every MaT. The disadvantage of this MaT is the visual aspect. The cabinets that are installed in the VIP jets have a high weighting factor in the appearance. That is why normally deco panels are installed in front of the structural panel. The added skin on the mortise would destroy the smooth surface on the corner of the cabinet. Therefor this concept has a lower ranking compared to the first three options.

Abbildung in dieser Leseprobe nicht enthalten

Figure 9: option 4 with T joint shape

2.3.5 Use-Value Analysis L Joint

To figure out the best way for the first test the use-value analysis has been evaluated (see Table 2. It will be evaluated under the same circumstances and requirements as the T joints. The description and requirements for the rating shall be looked up in Chapter 2.2.4.

The first option has the best overall results but is in none of the variables the alone leader. For the weight the second and fourth option are the worst because of the amount of skin left on the panel with the tenon or mortise. The first option is the winner when it comes to the weight of joints in a cabinet. Even though the difference between the first and the third option is small the long MaT got more skin surface and will probably need more adhesive for the bonding.

Abbildung in dieser Leseprobe nicht enthalten

Table 2: Use-Value Analysis for L Joints

The costs are not related to the skin surface and therefor the results differ a bit from the weight. The first and the third option will consume the least amount of adhesive. The correction of the problem with the lowered surface due to the shrinking adhesive (see Chapter 2.3.3) will also cost more. The surface is usually corrected with Aerok9111R[6].

For the manufacturability the fourth option is the best one. It has a mortise pocket and for that no lowering of the skin after the adhesive is cured. Additionally the pocket makes it easy to fill the adhesive in with pressure without putting aluminium tape around the edges of the tenon on the honeycomb. The second option has many disadvantages with the manufacturing phase. First it needs to be cored out after the CNC machine and loses the whole stiffness. Secondly it is difficult to clamp the big side of the tenon that it is bonded evenly to the mortise panel. The first and third options are similar in the way they are made. They have to be covered with aluminium tape on the side with the open honeycomb to provide the needed pressure for the adhesive to fill all the holes inside the mortise. The time exposure is evenly distributed except for the second option which has related to the manufacturability a bad rating. The reason for this is that the removing of the core will increase the manufacturing process of one cabinet dramatically (similar to Chapter 2.2.1). The last but for the first test most important factor is the strength. Option 4 has the advantage of the additional honeycomb skin on the transverse shear direction. This will get the allowable to a good value and might also increase the tension stress allowable a few pounds. The second option has the major advantage in strength with the big skin area that is bonded to the mortise panel. The bonded area is essential for a strong adhesive joint and therefor this technique would increase the strength greatly. Option 1 has a decent bonding area but can only use one of the tenon skin sides for bonding. Whether this is needed anyway has to be proofed and is probably not an issue. The third option will have too much of a depression on the skin that got lowered by the shrinking adhesive. This could lower the strength due to the changed shape of the skin. The first option shall be the choice for the first test coupon design and the company test. An additional advantage is that the L and the T coupon are very similar in the dimension and that makes it easy to manufacture it on the CNC and with the process specifications.

3 Test arrangement

3.1 Introduction

After evaluating the Option that is most appropriate for the panel joints when it comes to simplicity, manufacturability, weight, cost and strength, the test plan and arrangements are to be made.

The testing and by association the test equipment are located in the test facility of Bizjet. The test procedures and arrangements are designed to meet current regulations of the FAA. Based on the test equipment used for the evaluation of the panel pins allowables, which already ran through FAA controls and approvals, it is assured that the tests are performed in the appropriate environment.

The test plan is designed to meet the FAA regulations of 14 CFR shown in Table 3.

Abbildung in dieser Leseprobe nicht enthalten

Table 3: FAA regulations for test arrangement

These regulations are the determining factors for the certification and ensure test coupons with the potential of being an industrial standard. The tests are standardised to be comparable and assessable with different joint techniques.

3.2 Test equipment

The test setup seen in Figure 10 is designed to be interchangeable to every test in this project including tension, longitudinal shear, transverse shear and long beam bending.

Abbildung in dieser Leseprobe nicht enthalten

Figure 10: Test setup

The load actuator is hydraulically controlled and is operating in a deflection driven force. This is needed to display an accurate diagram with the load as the Y variable and the deflection as X variable. It is necessary that the load is applied at a constant rate so that failure occurs between 1 and 2 minutes. Also it is needed to optimize the results in regard to analysis and comparison. Therefore the load application also needs to be of a consistent rate between samples in the same batch. Additionally the Load actuator has to be capable of applying a load that is 30% greater than the anticipated failure load maximum of the test coupons to ensure that every single coupon will fail (ultimate load). For correct voltage values and, by implication, right measurements too the load cells and the displacement transducer are needed to be calibrated at least within a year prior to the test. The calibration must be executed in compliance with the NIST standards. The displacement transducer shall have the ability to measure 0.002 ± 0.001[10]. Further on the calibrated digital data recorder is capturing and recording the output from the load cell and the LVDT position transducer. Chucking the honeycomb panel would need a clamping force that is equal or higher than the anticipated max. failure load plus the additional 30% for the ultimate load. This force would extend the permissible stress for the stabilized core compression of the panel.

The following equation will show the needed clamping area visualized in Figure 11: Clamping area for the test panel:

Abbildung in dieser Leseprobe nicht enthalten

Figure 11: Clamping area for the test panel

The lowest Stabilized Core Compression value of the tested NORBOND™ panels has the “220-0155-500” with 300 psi[11].

Figure 11 shows the available clamping area for the panel. Since the whole width of the panel can be used, the only restrictive variable is the height for clamping. The maximum Force for Clamping (Cmax) can be calculated by using the max. failure load.

For the testing of the MaT coupons a max. failure load of 900 lbf is anticipated.

This anticipated value is 10% greater than the mean value of the panel pins in the testing.[12]

With the inclusion of the additional 30% we get an ultimate load:

Abbildung in dieser Leseprobe nicht enthalten

Figure 12: occurring Forces at the clamps shows the typical occurring forces in a setup with clamps.

Abbildung in dieser Leseprobe nicht enthalten

Figure 12: occurring Forces at the clamps

The Frictional Force (Ff) is the force which is working in opposite direction of the tension force (F). It is defined by the normal force and the friction coefficient:

Abbildung in dieser Leseprobe nicht enthalten

The coefficient of friction for steel in contact with a glass fibre epoxy composite can be compared with a carbon fibre epoxy composite[13].

Abbildung in dieser Leseprobe nicht enthalten

Additionally we need a safety factor to make sure that none of the failures is related to an in stabilized core:

Abbildung in dieser Leseprobe nicht enthalten

With the allowable for the Stabilized Core Compression, the width of the panel and the variable height we get the following equation:

Abbildung in dieser Leseprobe nicht enthalten

With that said the total height of the clamp needs to be about 5.22 inches.

The potential clamping area on the panel has a height of 4.7 inches and cannot be exceeded. Therefore a different method was evaluated with a pulling fixture consisting of seven pins which are applying the load in bearing stress. The load to failure on the seven pins with a diameter of 0.5 inches is well over the anticipated load to fail the joint. To prove this the following assumptions and equations need to be made: The failure off the panel due to the pins will either be caused by exceeding the shear allowable or the bearing stress allowable. For this reason both of the failures have to be checked.

The bearing stress can be calculated with the following equation:

Abbildung in dieser Leseprobe nicht enthalten

Dp is the diameter of the hole and has the value

Abbildung in dieser Leseprobe nicht enthalten

Whereas the thickness of the two skins is

Abbildung in dieser Leseprobe nicht enthalten

The expected load can be taken from the calculated ultimate force Fult from the clamping force.

If we split the force to the seven pins it leads to the following equations

Abbildung in dieser Leseprobe nicht enthalten

If it is conservatively assumed that the maximum bearing stress of fiberglass is 25% of the bearing stress for aluminium 2024-T3 (AMS-QQ-A-250/4)[14] e/D=1.5 then Fbru for fiberglass will be:

Abbildung in dieser Leseprobe nicht enthalten

According to the calculated occurring bearing stress and the allowable it has been proven to not fail with cause of bearing stress.

In the following the maximum pulling force for the shear allowable will be evaluated.

Figure 13 is showing the free-body-diagram for the shear forces caused by the pulled pin.

As is the shear area and consists of the length Ls multiplied by the thickness of the face sheet and is approximately ܣ௦ = [Abbildung in dieser Leseprobe nicht enthalten]

Figure 13: bearing stress reactions

The shear allowable Fsu will be calculated from the “HexForce_Technical_Fabrics_Handbook”[15].

With strength of 300[Abbildung in dieser Leseprobe nicht enthalten] and a thickness of 0.008in we achieve a shear allowable of:

Abbildung in dieser Leseprobe nicht enthalten

With the formula for the shear stress:

Abbildung in dieser Leseprobe nicht enthalten

The maximum load per pin is 3000lbf which is sufficient for the expected maximum load.

3.3 Test requirements

To design and plan the test as suitable as possible for optimized results, different requirements have to be fulfilled such as environmental and test object requirements. The numbers of coupons for the first tests are not based on establish a statistically FAA approved “A” and “B” basis allowable, but to constitute a mean value for the test results to avoid using numbers that are highly deviated from the true range of values. Therefore a number of three test coupons are selected to create a fairly precise result in order to analyse and improve the Joint. The requirements for “A” and “B” basis allowables will be discussed in Chapter 4.2.2.1.

Additionally the test coupons have to be tested in a certain environment to get comparable data with the panel pins. For realizing that, the data of the panel pin test was examined. The test facility had to have environmental standards of 65-85°F and 50-80% relative humidity. The load cell must have an output of 10 samples per second at a minimum and will be saved digitally. The A and B Basis shall be calculated in accordance with the statistical analysis of the MIL-HDBK-17G (See Chapter 4.2.2.1). The software tool STAT17F_EXCEL97-2003_Rev-4_2-8-07[16] is to be used. The loads shall be applied at a rate that failure occurs after 1-2 minutes. After the first test the rate can be adjusted, so that it failure in the right time range.

3.4 Testing procedure

The tests are divided in three different load directions. Figure 14: Definition of load directions shows all the different permutations that have to be tested. P stands for the load and the arrow indicates the direction. The directions can occur either from the varying clamping methods or from the immediate pulling direction.

Abbildung in dieser Leseprobe nicht enthalten

Figure 14: Definition of load directions

The following text describes the procedure for the tension test because it is the critical value for construction allowables. First of all the testing fixture has to be assembled for the tension setup. The date, time, humidity and temperature are needed to be recorded and to be mentioned on the result data sheet. They could affect the comparison of the data of different test runs. Each test coupon must be marked with a part number for identification. The specific code for the coupons is described in Chapter 4.2.1.3.

The test specimen shall be placed in the fixture with clamps as seen in Figure 15 and need to be verified as tight. The pneumatically driven clamps are designed to apply the right pressure on the test specimen in regard the core compression allowable of the sandwich panel. The load cell and position transducer with the deflection must be zeroed prior to beginning of the test. After start of the recording the load shall be applied to meet requirements in Chapter 3.3. The maximum load until failure has to be determined and recorded from the operator and/or the observing DER. For the company test no DER is needed because it is not yet FAA related. After the test the type or mode of failure needs to be determined and recorded. This is elemental for the development process and also important for quality insurance, to make sure that the test coupons were correctly manufactured.

Abbildung in dieser Leseprobe nicht enthalten

Figure 15: drawing of the test setup

The additional test that will be accomplished at the end of the project is the long beam bending test. This test is designed to proof the resistance of the new joint in regard to bending.

This method is a common way to show the allowables for sandwich structures because it is optimized in the dimension to get a value that is accurate and consistent. The long beam bending test is used to determine the face sheet properties[17]. The setup is shown in Figure 16 with the upper loading beam on top and the supports on the bottom of the test coupon. The maximum span is dimensioning every distance in this setup.

Abbildung in dieser Leseprobe nicht enthalten

Figure 16: Bending test setup

The test coupon with the length of 22 inches has to be placed in the test setup to ensure an evenly distance from each side to the deflection measurement. This is needed to achieve the desired values. After that the upper loading beam must be placed on the test coupon with a little pressure to prevent it from sliding away. The four points are mounted with a rotational degree of freedom to ensure that the force of the upper loading beam has always a contact pressure that is within the range of core compression of the panel. This rotation of the supports can be seen in Figure 70. The upper loading beam will be increasing the pressure on the panel until a failure can be noticed.

4 Test phase

4.1 Introduction

The purpose of this chapter is to ensure that the appropriate test coupons are designed and manufactured. Additionally the progress of this project with the aim to create a more efficient joint with MaT than with panel pins has to be demonstrated and followed. After evaluating the best joint technique and the compatible test setup the test coupons shall be designed. The test arrangement as shown in Chapter 3 has dimensions that are fixed and requires test coupons that are designed to fit these dimensions. The test specimens will be manufactured at BizJet International. Conformity inspection of the test coupons will be done before testing to ensure the implementation of the process specification for manufacturing the coupons. The process specification will be on the first page of each drawing and are instructions for the manufacturer to avoid mistakes and create conformity. The main focus of the tests is on tension because it is the critical factor for a valuable allowable. The shear loads are predicted to be very strong just by using the MaT to block the movements in the shear directions. Visualized in Figure 18 and Figure 18 the failures of the shear tests are predictable and valued as high enough to be comparable with panel pins. For this reason the first test coupons will be tested in tension.

Abbildung in dieser Leseprobe nicht enthalten

Figure 18: TLS points of failure Figure 18: TTS points of failure

The project plan seen in Figure 19 was designed to visualize the different steps that are needed and used for this research and development project. The first step in this project is to evaluate the needed requirements. These steps include requirements from the test and the basic preparation for that through to the panel pin data that is the basis for comparison. Different materials were evaluated and checked for their performance in the existing load cases. After gathering all information and creating the test coupons the company test has to be done. The first test is to get data for different configuration to compare and elaborate the best method to continue the project. It is not expected to be better than panel pins after the first test run. After analysing the test results the value analysis shall use the data to create a more efficient way to build up the MaT and maybe an even stronger joint. With the goal to be more efficient than panel pins in at least 3 of the 4 criteria such as weight, strength, manufacturability and time exposure the steps of improving the MaT joint and testing will be repeated until the goal has been reached.

Abbildung in dieser Leseprobe nicht enthalten

Figure 19: Project plan

The following chapters are organized in an order to show the progress of this project. The first test run is the most important and the most difficult because the configurations have to be chosen in a way to show the best comparison data for further improvements. It is also the “go or no go” company test which will decide the on-going steps in this project and the amount of money that will be invested to develop the MaT joint.

4.2 First test run

4.2.1 Test coupons

4.2.1.1 Design

The tests are part of a research and development project and each test will change the design of the next test coupons to improve it in strength, weight and manufacturability. The first test will include 42 test coupons. Subdivided into different configurations the test shall show the strongest joint. Each configuration contains three test coupons. This ensures the accuracy of the test result, the right manufacturing and shows the level of consistency. The shape of the tenon shown in Figure 20 is the same for the L and for the T joint and will not change over the entire project. The undercut is an often needed method in CNC cut parts to provide a proper fit between the tenon and the mortise.

Abbildung in dieser Leseprobe nicht enthalten

Figure 20: Dimensions for the tenon

At the same time the undercut got a good transmission of force whereas a sharp edge would dramatically increase the probability of cracking. The height of the tenon is limited by the second panel with the mortise. It has to be 0.1 inch shorter than the thickness of the second panel. This shortening is provided to allow enough space between the skin on the lower side of the mortise and the skin of the tenon. It has to be ensured that the skin [1] seen in Figure 20 fits tight to the other panel.

The diameter of 0.125 inches is the smallest tool size Bizjet has in stock for the CNC machine and is therefore the chosen size for the undercut. The length of the tenon is two times the width of the panel with the mortise in it (0.5 inch panel = 1 inch tenon). This will guarantee the best strength properties due to the use of maximum surface area for the bonding and additional skin area for the transverse shear (see Figure 18: TTS points of failure). To get additional surface area for the adhesive the Nomex core has been removed to a certain length inside of the tenon. For the first test runs the core of the tenon will be removed by 1/8 inch (see [2] in Figure 20). This should provide enough bonding surface for the tenon. It also allows the adhesive to flow around inside the MaT without obstacles. 1/8 Inch of removed core material complies with the length of one core hexagon[18] (See Figure 21).

Abbildung in dieser Leseprobe nicht enthalten

Figure 21: tenon with 1/8 inch removed core

The mortise has two different configurations for the removal of the core to show the effects of this parameter in regard of the panel joint strength. Also the weight will be a critical factor for the cored out mortise because the weight of the injected adhesive might double with 1/4 inch instead of 1/8 inch core out. The better configuration will be evaluated in Chapter 4.2.2.

4.2.1.2 Adhesive

The MaT joint is an adhesive bonding and therefor the appropriate adhesive has to be provided to ensure the best mechanical qualities. In this project three different adhesives will be considered. Each adhesive has strength and weaknesses in regard to shear strength, peel strength, tensile strength, viscosity and processing. The data for the analysis goes from data sheets of the vendors or other sources to the experience of the companies’ employees. ATR 525 is a rapid curing two-part structural adhesive which is specifically designed for the “Panel Pin Honeycomb Panel Fastening System” and is produced by AAR composites. It has the best performance when used for bonding an epoxy or plastic material. The ATR 525 is already being used by BizJet and is supplied with a pressurized gun for injecting the adhesive into the panel pins. This gun can be used for the MaT pocket to fill it with pressure and provide an even distribution of the adhesive. The pot life of ATR 525 is 15 to 30 minutes (depending on the mass) once it is mixed and shall not be exceeded. The mixing ratio is shown in Table 4 and consists of a white paste (part A) and a black paste (part B).

Abbildung in dieser Leseprobe nicht enthalten

Table 4: Mixing ratio of ATR525

The second adhesive is the two component low density adhesive/filler Aerok 9110R[19]. This adhesive is developed for composite and aluminum honeycomb panels. Often used for “ditch and pot/cut and fold” applications and as edge filler. According to the vendor it will meet the requirements of FAR/JAR/CS 25.853 (a) - 60 second vertical burn and ABD 0031/D6-51377 Toxic gas emission.[20] The two components must be mixed in the ratio shown in Table 5 to maximize the structural properties. Aerok 9110R is available in a pressurized gun to apply the adhesive into the mortise without having to mix it by hand. Unfortunately Bizjet has this adhesive kit not in stock so it has to be hand mixed and then filled in a gun. Aerok 9110R also has a very low density with 0.7 ݃/ܿ݉ଷ. Hysol9309.3NA and ATR525 have a density of about 1.1 ݃/ܿ݉ଷ. This adhesive would be preferable in the aviation to save additional weight.

Abbildung in dieser Leseprobe nicht enthalten

Table 5: Mix ratio for Aerok 9110R

The last option that will be presented is by experience of the cabinet makers at BizJet the strongest adhesive. Hysol EA 9309.3NA[21] is a two-part paste adhesive and has very good bond line control. Very strong in bonding honeycomb core with high shear and peel strength. It is resistant to humidity water and most common fluids. Hysol 9309.3NA comes in two parts and cannot be mixed in an outlet of a pressurized gun. The part B has only a viscosity of 0.2 Poise and is therefore not squeezable through a tube. This means that this adhesive has to be hand mixed (see Figure 22) in the right ratio by weight shown in Table 6: Mixing ratio for Hysol 9309.3NA. The red paste is the part A and is very sticky and gets more and more fluid by mixing the blue colored part B.

Abbildung in dieser Leseprobe nicht enthalten

Figure 22: Hysol parts in a container Table 6: Mixing ratio for Hysol 9309.3NA

The pot life of Hysol 9309.3NA is 35 minutes with a mass of approximately 450 grams. It is highly recommended not to mix quantities over 450 grams. The buildup of heat is normal due to an exothermic reaction but a mass over 450 grams can cause uncontrolled decomposition of the Hysol.

The most important properties of the adhesives that are used in the MaT joint are strength and the handling for the manufacturers. The structural properties are shown in Table 7 to compare each adhesive for the best qualities.

Abbildung in dieser Leseprobe nicht enthalten

Figure 23: Tension

Abbildung in dieser Leseprobe nicht enthalten

Figure 25: Lap shear Figure 24: Compressive

Figure 23[22], Figure 24[23] and Figure 25[24] visualize the different load directions and how they affect the adhesive. To ascertain the appropriate adhesive for the MaT joint the appearing forces shall be evaluated. Figure 26 shows the forces that are occurring when the MaT joint is tested in the tension setup. The bottom of the tenon has tension forces and the sides of the tenon are similar to lap shear forces. The sides of the tenon without skin will not have a big impact on the tension allowable because it is bonded core to core. Therefore the MaT joint has more shear area than tension area and the distribution is about 75%-25%. Hysol 9309.3NA has very good shear resistance and could be a very good choice for bonding the MaT. The tension strength of ATR525 and Aerok 9110R is higher and could be also a good choice for the MaT joint because they are easier to mix together or do not even have to be mixed by hand. Although this test is focused on strength the economic impact is always considered. For this reason the advantage of the stronger shear strength of Hysol will probably not compensate the higher price for each future joint. Additionally, the tension area for the L joints will be bigger than the shear area because the side on the outside honeycomb cannot be bonded.

The first coupons will therefore be tested with the configuration of ATR525 and Aerok 9110R. These tests will show the compatibility with the adhesives to the MaT joint and if the values given by the vendors are useful for this joint. The values of the vendors refer to an optimized bonding with a very thin layer between two adequate surfaces. In this joint the surfaces for bonding are small and will be hard to control.

Abbildung in dieser Leseprobe nicht enthalten

Table 7: structural properties of the adhesives in comparison with the costs[25]

Figure 26: Occurring forces in the T Joint during the tension test

4.2.1.3 Requirements

After discussing the design and the appropriate adhesives which are significant to a successful joint this chapter defines the requirements for the test coupons such as the specific code for each test coupon, dimensions and tolerances. These requirements are needed to provide uniformity in the manufacturing and testing process. The specific code must give every test coupon a unique identification and preferably describe the configuration of the labeled sample.

The configuration for the first test differs in the properties shown in Table 8:

Abbildung in dieser Leseprobe nicht enthalten

Table 8: configurations for the

The code for the coupons characterizes the configuration completely and makes it easy to analyze it after the test. The first part of the code is the Joint. Since this project includes two joints the indication is either “-L or -T”. The second digits of the code are describing the amount of core that has been removed. Two different configurations will either be “-18 for 1/8 inch or -14 for 1/4 inch”. The third part of the code indicates the group of samples that are composed of the same configuration. This simply adds up on each group for example “-001 or -005”. The fourth digits will show the order of the coupons in which they got tested in the group. The last part of the code indicates the used adhesive. For the first test it will only be ATR525 and Aerok9110R. A complete code for a test coupon that is an L joint, has 1/4 inch core out, is the first test run from group 5 and bonded with ATR525 will have the code: “L-14-005-1-ATR”.

The second requirement for the test coupons are the dimensions. The test arrangement explained in Chapter 3 has a fixed setup and the test coupons have to meet the dimensions for this setup. The drawings with the dimensions are to find in Attachment F.

Each coupon that is tested on tension has 7 holes with a diameter of .531 inches on the panel with the tenon. These holes ensure a constant stress distribution over the panel and a vertical pull.

For everything that will be manufactured it is needed to state the tolerances. The test coupons are being manufactured on a CNC machine and for that the tolerances can be picked very small. This is especially needed for the width of the mortise because the panel has to fit into that very tight to ensure an easy manufacturing on overhead installations. The tenon shall not stick loose in the mortise. Overall dimensions are will have tolerances according to the Bizjet standard tolerances[30] shown in Table 9.

Abbildung in dieser Leseprobe nicht enthalten

Table 9: Tolerances in the drawings

4.2.1.4 Manufacturing

After collecting the needed data for the test coupons and preparing the drawings for the manufacturer, it is now the milestone of getting the sandwich panels cut on the CNC machine and bond them together after a certain process specification. Process specifications are needed as a communication between the engineer and the manufacturer to ensure the right methods and tools are used for the test coupons. Additionally it creates conformity which is needed for a successful test result. The smallest nonconformance can create a higher standard deviation in a test group and could lower the allowable of the MaT joint. How the allowables are calculated is described in Chapter 4.2.2.1. The following text describes the process of manufacturing the test coupons for the first test.

The first step is to clean the side edges of the panel with the tenon. For this a razorblade is needed to ensure an even core surface on the side edges. Next, the core of the panels has to be removed depending on the dedicated configuration. The tool to remove the core can be either a dremel or a die grinder. The preferred tool is the die grinder for the easier handling and more accurate results. 1/8 inch of removed core correlates to one hexagon of Nomex core. Since this procedure cannot be 100% accurate this visualization helps the manufacturer to get a sufficient result. An air duster/blow gun shall be used to clean the mortise from the removed core. A table on the drawing shall be provided to supply the manufacturer with the needed information such as material, quantity and details for the notes on the drawing. Table 10 is representing an example for the information that is included in this table.

Abbildung in dieser Leseprobe nicht enthalten

Table 10: information table on the drawing

It is required that the test coupon panels are hold in a perpendicular position for guaranteeing the right pulling direction. Different methods can be used to ensure a perpendicular position such as screws, clamps or super glue. To hold the panels together with pressure the technique of screwing is used. After the mortise and the tenon got stuck together and hold in a perpendicular position there are four holes to access the mortise with the adhesive. To ensure a complete fill of the mortise only two holes are needed. Figure 27 shows the drawing view of the mortise and it is visible that the corners with a flag note 1 are .05 inches longer than the other. The additional depth is designed for the possibility to inject the adhesive with the nozzle sealing against the tab. Flag note 1 is representing following process specification:

“Inject mixed adhesive into one of the holes while applying firm pressure to the adhesive gun so the nozzle seals against the tab. Continue filling the cavity surrounding the mortise until the adhesive oozes out of the opposite vent hole. Place nozzle over opposite vent hole and inject adhesive until it oozes out of the first hole used for injecting.”

The open holes for the adhesive injection are visualized in Figure 28. To guarantee a pressurized injection to fill all the gaps inside the joint, aluminum tape shall be bonded around the edges of the MaT. Holes for the injection must be provided in the aluminum tape.

Abbildung in dieser Leseprobe nicht enthalten

Figure 27: mortise of the first test coupons

After the adhesive is successfully injected according to the process specifications the aluminum tape must be removed and the edges shall be cleaned to assure a smooth surface. Shortly before the test the samples then need to stay in the heating room at 120° Fahrenheit. This guarantees a shorter cure time of 6-8 hours. At normal room temperature of 72° Fahrenheit a full cure would need about 5-7 days.

Abbildung in dieser Leseprobe nicht enthalten

Figure 28: Aluminum tape for the injection

4.2.1.5 Strength

The first test is all about the qualities of the MaT when it comes to strength in tension. The strength allowable is currently the critical part of designing a cabinet in regard of structural reliability. Also the first test shall show the advantages and disadvantages of every configuration that is possible even though some of them can only be assumed. This is the foundation of further tests and development. The values for tension, longitudinal shear and transverse shear are compared with the panel pin mean values and allowables which originated out of the previous tests. The values shown in Table 11 are mean values of twelve test coupons for each configuration. In the MaT test only consists of three test coupons for each configuration. This will ensure not a perfect but a sufficient quantity for comparison with the panel pins.

Abbildung in dieser Leseprobe nicht enthalten

Table 11: Panel pin mean values of the tension test

The longitudinal and transverse shear strength is expected to be good enough due to the tenon that will block in the mortise for acceptable results. The testing for them will be performed in the final test for the project with the intended tension properties.

4.2.2 Evaluation of the test results

4.2.2.1 Standard deviation and allowables

The main part of this project is the first test run and the analysis of the results. After determining all requirements that are needed to accomplish a successful test and manufacturing test coupons after certain procedures and to fulfil the needed qualities, the first test is done and will be evaluated. Before the results will be analysed this chapter will describe how the allowables will calculated and what is important to achieve valuable allowables. The first variable for the allowables is the standard deviation.

The standard deviation is an equation in statistics to measure the range of the values from the average. A lower standard deviation indicates that the data points of the measured area are close together whereas a high standard deviation indicates that the failure loads are spread out over a large area of data points. The following equation[31] is showing how the sample standard deviation is calculated:

where xi are the tested failure loads and [ LV the mean of the tested group, while N is the quantity of tested samples. The standard deviation is an important factor for calculating the A and the B basis allowables for the MaT joint. The A and B basis are statistically- based material properties. The A basis defines a 95% confidence that 99% of the data will fall above the allowable. The B basis defines a 95% confidence that 90% of the data will fall above the allowable. The evaluation of the less restrictive B-basis allowable is shown in Figure 29[32]. The B-basis allowable is the 10th percentile of a material regarding to the properties.

However it has to be ensured that the B-basis allowable is also the 95th percentile out of a random test of n specimens because the true population is not known. Therefore the B-basis is statistically based on a random sample with n specimens, such that a repeatedly test of the specimens and calculating of the B-basis values will cause 95% of them to fall under the desired B-basis allowable.

Abbildung in dieser Leseprobe nicht enthalten

Figure 29: B-basis allowable

The A-basis has the same calculation method, just with the 1st percentile of the true population. The Composite-Materials-Handbook-Mil-HDBK-17 Volume 1 describes three different methods to calculate the right allowable. The “Normal Distribution Statistics”, the “Lognormal Distribution Statistics” and the “Weibull Distribution Statistics” are the existing methods to calculate the A- and B-basis allowables[33]. Which method is used will be decided after an observed significance level (OSL). The method with an OSL higher than 0.05 is the choice for calculating the A- and B-basis allowable. As an example of how to calculate the B-basis allowable the procedure for a normal distribution will be described in the following text.

The B-basis can be calculated with the equation:

Abbildung in dieser Leseprobe nicht enthalten

With B as the B-basis allowable, [ DV WKH PHDQ RI WKH WHVWHG JURXS DQG V DV WKH VWDQGDUG deviation. The factor k for a B allowable (90% coverage and 95% confidence) can be selected from Table 12. This table shows why for the MaT company test each group has three test coupons with the same configuration because the step from 2 to 3 samples has a big impact on the k factor. For a Test that is designed for FAA approval it is recommended to test 12 or more coupons to reduce the k factor as much as possible but also not to design it needlessly expensive. That is why BizJet has tested 12 coupons for the Panel Pins in every configuration.

Abbildung in dieser Leseprobe nicht enthalten

Table 12: Normal Dist. Factor k for a test group of n<16

The methods for the A- and B-basis allowable can be looked up in the “Composite Material Handbook” from the Department of Defense. With designing a cabinet it is always recommended to choose the A-basis allowable for stress analysis. Although it is probably not easy to succeed, it could be easier than the proof of a save design with the B-basis allowable. B-basis strength is allowed in redundant structure in which failure leads to a different safe load distribution.

4.2.2.2 Test results

The appropriate test coupons have been designed and manufactured after certain process specifications. The requirements for the test setup and the samples have been made to achieve optimal values. All in all the first test had satisfying results. The aspiration to gain tension strength with results close or better than panel pins was successful. Even though some configurations are only about 2/3 as strong as the panel pins there are configurations with 15% higher values. Weak and strong joints are the key in this project for further improvements. Closer analysis will bring clarity about the different failure modes that occurred. In the following text each specific configuration will be compared and discussed with the equivalent.

The first configurations that shall be discussed are the properties of the adhesives in the MaT joint. For that we had Aerok 9110R and ATR 525 as the compared adhesives. With analysing the results shown in Table 1 it is obvious that ATR525 is providing less strength than Aerok 9110R. This configuration of 1/2 panels and ATR525 was the weakest with only 2/3 of the panel pin strength. To verify this statement these two configurations are just an example. The complete test result data sheet can be examined in Attachment G. However, in each comparison Aerok 9110R had the better tension strength.

Abbildung in dieser Leseprobe nicht enthalten

Table 13: Configuration with two adhesives

For the consistency of the different adhesives it can be evaluated, that the standard deviation of each ATR525 test configuration is lower and therefore better than the Aerok 9110R samples. The standard deviation of the ATR test coupons is Satr=24.5 whereas the Aerok test coupons are tested with a standard deviation of Saerok=38.85. As described in Chapter 4.2.2.1 not only the failure load is important for the test coupons but also the standard deviation is a critical factor to accomplish a valuable allowable. The failure mode in both test coupons was adhesive failure. This means that Aerok 9110R has better resistance against tension in the adhesive. An additional bad property that is visible in Figure 30 and Figure 312 is the uneven distribution of the adhesive in the MaT. Even though the plan to inject the adhesive with pressure to ensure that all gaps are filled, air pockets are visible in the Aerok MaT. Additionally the injected adhesive tends to squeeze out of the MaT inside the panel which can lead to nonconformity, interference and even issues with flammability.

Abbildung in dieser Leseprobe nicht enthalten

Figure 30: ATR sample -003

Abbildung in dieser Leseprobe nicht enthalten

Figure 31: AEROK sample -001

The following diagrams of the ATR525 and Aerok 9110R samples are similar in the characteristic. Therefore only one diagram shall be described. The recorded data of the calibrated digital data recorder is visualized in Figure 32 and Figure 33.

Abbildung in dieser Leseprobe nicht enthalten

Figure 32: Data of sample -001 in a diagram

The diagrams of the tests provide useful data and information about the failure modes. The different occurrences can be divided into different deflection zones:

0-0.2 inches: The load actuator is building up more and more stress in the MaT with higher deflection. The small drops in the load are small parts of the adhesive that are already failing.

0.23-0.27 inches: The maximum of tension load has been reached. The failure mode for that is the crack of adhesive in the MaT. The load is dropping and increasing again because it is held up by the skin that is now withstanding the stress.

0.29-0.30 inches: The last failure mode of the joint is the skin failure and has still good strength properties. After the skin has failed the load is dropping to a minimum and the test is finished.

The only anomaly with the first test run of the ATR sample is the triple failure. It is assumed that the squeezed adhesive into the panel is the first point of failure. After that the adhesive broke inside the panel and the last failure is the skin.

Abbildung in dieser Leseprobe nicht enthalten

Figure 33: Data of sample -003 in a diagram

The second comparison shall be the test coupons with 1/8 inch and 1/4 inch removed core. The reason why two different configurations were tested is to evaluate the impact of the amount of adhesive inside the mortise. The configuration without any removed core was not taken into account because it had to be ensured that the adhesive fills out the whole mortise.

Abbildung in dieser Leseprobe nicht enthalten

Table 14: Configuration with different sizes of removed core around the mortise

For this comparison the panels are ¾ inch thick and therefore have a higher failure load. Nevertheless the difference between the two sizes of the removed core around the mortise is similar in every configuration (see Attachment G). The 1/4 inch removed core is about 10-20% stronger than the 1/8 inch. The issue is that the additional weight for each joint nearly doubles with the additional 1/8 inch of removed core around the mortise. Whether the additional weight per joint is worth the stronger joint will be discussed. A possible solution could be to write in the process specification that each joint calls out whether it has 1/8 or 1/4 inch removed core. This depends on the design of the cabinet and if a joint needs enforcements due to stress distribution or not. The failure modes of both configurations are the same as seen in Figure 34 and Figure 35. The first failure on each sample was the adhesive that broke on the bottom of the skin. After that the skin on the short side of the mortise had not enough area to hold against the forces that have been already applied.

Abbildung in dieser Leseprobe nicht enthalten

Figure 35: Aerok sample -009 with 1/8 inch Figure 34: Aerok sample -008 with 1/4 inch

The diagrams in Figure 36 and Figure 37 are showing the occurrences that are visible in Figure 34 and Figure 35. At a deflection of approximately 0.2 inches both samples got the first failure.

Abbildung in dieser Leseprobe nicht enthalten

Figure 36: Data of sample -008 in a diagram

Whereas the test coupon -009 has already reached the maximum load and drops down to a minimum, the sample -008 is recovering from the short load drop and is increasing the strength to a higher maximum. The critical point of failure for this test coupon is the skin around the mortise. Due to the high amount of Aerok 9110R inside the mortise the attached skin has a bigger area. In the following the diagrams are going to be described:

0-§ LQFKHV the load actuator is applying the load with a constant increase until the first point of failure is reached

§ inches: The first failure of the adhesive is causing the first decline of the load.

0.2-0.3 inches: The sample with 1/4 inch removed core is increasing load again over the maximum of the first failure whereas the 1/8 inch sample is sinking below 200 lb.

0.3-1 inches: The skin fails in the 1/4 inch test coupon and decreases the applied load dramatically to a low value. The further deflection is unimportant for this test. The intention is merely to separate the two panels.

Abbildung in dieser Leseprobe nicht enthalten

Figure 37: Data of sample -009 in a diagram

The analysis of the two configurations with different amount of removed core showed that the 1/4 inch test coupon uses the skin to get a high maximum failure load. This shows the good shear properties of Aerok 9110R to resist breaking on the side of the mortise. Nevertheless the bigger panels contain more air pockets due to imperfect filling. This could lead to high standard deviations and therefore lower allowables.

The last analysis will be between the L and the T joints in regard to the panel pins. Table 15 is showing the comparison for the same configurations (panel size and removed core) of L and T joints. Overall the L joints are stronger compared to the panel pins than the T joints.

Abbildung in dieser Leseprobe nicht enthalten

Table 15: Comparison L and T joints

It is assumed that the MaT compared with the Panel Pin got a bigger surface that the panel with the tenon is pulling. The reason why the MaT has a higher resistance to tension can be speculated with Figure 38. The test coupon has not only tensile forces but also a moment and therefore a bending in the panel. What can also be observed is that the core next to the mortise fails first due to compression loads that occur during the bending. This influence is a major issue and raises the question whether the test setup is right for the L joints. For a tension test the first failure mode should not be the core compression due to bending.

Figure 38: 1/2 thick L joint with Aerok 9110R before failure

The analysis of the diagram will show the different failure modes. The three test runs of the 18_-013_9110R_500L are shown in Figure 39. Even though the different graphs are shifted in the x axis the characteristics are similar.

Abbildung in dieser Leseprobe nicht enthalten

Figure 39: Data of sample -013 in a diagram

0-0.15 inches: The load is increasing and the panel begins to bend.

0.15-0.2 inches: The bending of the panel is forcing the core to compress due to high core compression loads. The load recovers fast and begins to increase again.

0.4 inches: The maximum tension load has been reached and the test coupon has the appearance as shown in Figure 38. The failure at maximal strength is core and adhesive failure.

The L joint coupons with 3/4 inch thick panels do not have that critical bending of the panel due to the higher stiffness. Nevertheless the core in the test coupons also failed due to compression (see Figure 40).

Figure 40: 3/4 thick L joint with Aerok 9110R before failure

After analysing the different configuration with useful results the next step is to analyse unusual samples. Therefore the next two test coupons that are worth analysing can be described as outliers. What makes this a good evaluation is that they are outliers with surprisingly high tension values. Outliers are test result values that are outside other values and are normally reviewed as a non-representative value. In tests for an evaluation of allowables the outliers normally increase the standard deviation and therefore lower the allowable. Alternatively in research and development projects the outliers can help in the analysis for possible weaknesses or strength. The samples two and three were the first to be manufactured and due to mistakes of the manufacturer the panels were not correctly hold in a tight position. This allowed the adhesive to squeeze easily out of the mortise and under the skin of the side edges (see Figure 41 and Figure 42). In consequence of that the adhesive could bond not only from the lower side of the skin but also from the upper side. The triple failure of the test coupon with core, skin and adhesive is a sign for a promising technique.

Abbildung in dieser Leseprobe nicht enthalten

Figure 43: Data of sample -010 in a diagram

The major issue with bonding the side edges will be flammability. The discussion with the associated engineer for flammability has resulted in a positive conclusion. The appropriate authority FAA published a new policy statement for the “Flammability Testing of Interior Materials”[34] implying that the acceptable methods of compliance with flammability requirements have changed.

The policy statement is stating the following for bonded joints such as Ditch and pot, Cut and fold, Tab and slot, Mortise and tenon, T-joints, Bonded pins:

“Compliance of a bonded joint construction can be shown by:

OPTION #1: similarity to the base panel when the following are met:

1) The adhesive is an epoxy-based material ݱ

2) The panel is a honeycomb core panel with

composite skins meeting § 25.853(a), appendix F, part 1 (a)(1)(i), 60-second VBB, which is the ݱ

3) Compliance data used for similarity analysis.

The exposed adhesive is inside the bent/joined panel (e.g., inside cut) ” [35] ݱ

Since the two panels are tightly connected and the adhesive would be inside the joint respectively under the skin, the MaT joint would not need to be checked for flammability. This allows adhesives such as ATR525 and Hysol 9309.3NA which are not produce to meet the requirements of FAR/JAR/CS 25.853 (a) - 60 second vertical burn. The achieved acquirements are the basis for the ongoing improvements and the second test. A secondary goal is to lower the average weight of 30g per joint which would double the weight of a panel pin joint. The other diagrams are attached and can be found in Attachment I.

4.3 Second test run

4.3.1 Improvements

The first test is completed and brought useful results for further development tests. The second test run can be considered as a research test. It is not expected to be the final result but shall include the perceptions that are made on the outliers of the first test. These test coupons shall show the effect of the bonded side edges for the tension strength.

The design of the test coupons has changed in terms of the mortise and the way of injecting the adhesive. Analysis of the first tests have also shown that the skin around the mortise tend to break on the edges of the dog bone shape seen in Figure 44 and Figure 27. This is expected but probably not the best design to achieve maximum strength. The new design of the mortise shall minimize the amount of critical corners to maximize the strength of the skin.

Abbildung in dieser Leseprobe nicht enthalten

Figure 44: Common Failure on the mortise skin Figure 45: New mortise design

The mortise shape with rounded corners seen in Figure 45 has beside the structural advantage also the good property to leave no mortise visible after the panel with the tenon is installed. This prevents the needed closure of open holes which was needed on the first test coupons. The injection of the adhesive into the MaT is enabled with the hole on the bottom side of the mortise.

Abbildung in dieser Leseprobe nicht enthalten

Figure 46: Injection hole of the second test coupon

Abbildung in dieser Leseprobe nicht enthalten

Figure 46 shows the new test coupons with the open hole for injection. Flag note 1 describes the process specification with the following instructions:

“ Inject mixed Aerok 9110R into the hole while applying firm pressure to the adhesive gun so the nozzle seals against the tab. Continue filling the cavity surrounding the mortise until the adhesive oozes out of the opposite gaps.”

The gaps between the mortise and the tenon are very small which means that the adhesive has a lot of pressure on both sides. This should create less air pockets inside the MaT to prevent large standard deviations. The side edges are bonded with Hysol9309.3NA because it has the best bond line control which is needed to get the 0.02 inch thick face sheet bonded strongly to the panel. Additionally the panel surface has to be treated to optimize the properties for bonding. Table 17 is explaining the surface preparation for a bonding with Hysol9309.3NA in reference to the “Hysol Preparation Guide” by Loctite. The explosive view of the test coupon shown in Figure 47 is illustrating the new bond line that characterizes the second test coupon. The adhesive will be brushed evenly on the surface shown in the figure.

Abbildung in dieser Leseprobe nicht enthalten

Table 17: Hysol surface treatment[36]

Abbildung in dieser Leseprobe nicht enthalten

Figure 47: Bond line on the side edges

The treatment of the surface before the bonding is very important and should have a process specification to prevent unconformity and therefore not constant strength properties. Another major change is the removal of the core inside the mortise which is avoided for the second test coupons. The tenon though has still 1/8 inch of removed core to secure the bonding of the skin that is inside of the tenon. For this test the weakest panel of the first test (T joint on 1/2 panel) is picked to improve the properties where it is most needed.

4.3.2 Evaluation of the test results

The results of the second test run are satisfying in both, strength values and the deviation, which offers the capability for a proper analysis. Although a small standard deviation is wanted for the final test, this test is only to improve the test coupon in regard to strength. The test coupons are deviated with a pitch of 80 lbf and the first and third sample are deviated by 180 lbf which is about 30% difference. A large deviation like that raises the question what property of the test coupons led to it. The resulting data of the second test is summarized in Table 18 and is showing the weak sample (red marked) and the strong sample (green marked). The mean value of this test was 713lbf and is 16% higher than the panel pins whereby the strong sample had a 30% and the weakest a 3% advantage. These values led to a standard deviation of 82 which is rather high. Although the standard deviation of the panel pins for these ½ panels (T joint) was 151, the aim for a company test of the MaT shall be SMaT = <50.

Abbildung in dieser Leseprobe nicht enthalten

Table 18: Configuration with bonded side edges

The first perception that can be used out of this test that it is not needed to remove the core of the mortise to enhance the strength. Quite the contrary, it is strong although many work hours are going to be saved by not removing the core. For the next test it is also considered to leave the tenon in its original shape. The second perception is seen in Figure 48 and Figure 49 which illustrate the failed test coupons. As seen in the pictures the weaker sample has a different appearance around the MaT. The Aerok9110R (light pink color) got squeezed out of the MaT in both samples but the weaker joint has more Aerok9110R on the side edges. Since the test coupon with more Hysol9309.3NA is stronger than with Aerok9110R the next test coupons will only consist of Hysol9309.3NA.

Abbildung in dieser Leseprobe nicht enthalten

Figure 49: weakest sample of the second test run Figure 48: strongest sample of the second test run

The diagram in Figure 50 is characterizing the joint as consistent in its behaviour. Even though the deviations are high, the points of failure are the same. The detailed analysis of the diagram will be explained in the following text:

0-0.18 inches: The load is increasing and the panel is stiffer than in the first test due to the additional bond line on the side edges.

0.18 inch: All test coupons are failing due to a skin and core failure on the side edges. The side which fails first is determined by the amount of Aerok9110R that is squeezed underneath the side edges.

0.18-0.28 inches: The second and the first test run are dropping immediately to approximately 300 lbf. The entire side edge has been peeled off. The third test run is peeling off the side edge in 3 steps. First step is one side of the skins and then the other side. The last drop is caused by the side edge near the MaT that is holding the tension load for quite a while.

0.29-0.3 inches: The repeating increase of the tension load is caused by the MaT that are ripped out of the skin. This failure mode is in skin and core too.

Summarizing it can be said that the second test run was a success because it provided useful data for the final test such as the difficulty with two adhesives that are applied with different methods. The injected adhesive has more pressure than the brushed adhesive and therefore pushes it away.

Abbildung in dieser Leseprobe nicht enthalten

Figure 50: Data of sample -015 in a diagram

4.4 Third test run

4.4.1 Improvements

The perceptions that are made in the second and first test run are now to be incorporated in the last and final test coupons. These test coupons shall fulfil the needed requirements to end this project described in Chapter 4.1. Therefore the steps of manufacturing are illustrated step by step in the next chapter. The third test run not only includes tension tests but also investigates the strength of transverse shear and longitudinal shear values. Whether the technique which is optimized for tension tests also is effective for the other two load directions will be determined in Chapter 4.4.3. The first major improvement for the third test coupons is the single use of Hysol9309.3NA as an adhesive. With the good bond line control and the results of the second test in which the test coupon got weaker with less Hysol9309.3NA the choice is inevitable. This adhesive allows good bonding with a minimum amount of adhesive. This will make the joint lighter and stronger at the same time. The second major change is that neither the core of the tenon nor the core of the mortise gets removed. This will accelerate the manufacturing process and saves weight without an unnecessary amount of adhesive. The last change includes he filling of the MaT without pressure. This avoids holes and gaps in the joint that are needed to be filled after the joint is merged. The problem of a flowing adhesive in an overhead installation will not be an issue because the small amount of Hysol9309.3NA will stick on the skin and core instead of dripping off to the ground.

4.4.2 Manufacturing

The first step in the manufacturing process is to detach the expendable core that is longer than the skin. For this most commonly a sharp razorblade is used to cut the core perpendicular to the honeycomb (see Figure 51).

Abbildung in dieser Leseprobe nicht enthalten

Figure 51: Trimming of the core

After correcting the edges of the panels it is now important to prepare the bonding surface properly. The method for the surface preparation is already described in Chapter

4.3.1 and concerns the surfaces shown in Figure 52. To roughen the surface it is recommended to use fine-grit emery paper shown in Figure 53.

Abbildung in dieser Leseprobe nicht enthalten

Figure 53: Sand the surface with sandpaper

Figure 52: Surface preparation area

After the surface is properly prepared the area has to be cleaned with acetone illustrated in Figure 54. The mortise shall have no lose core or dust inside to prevent impurities that will weaken the bonding.

Abbildung in dieser Leseprobe nicht enthalten

Figure 54: Cleaning the surface with acetone or MEK

After preparing the surface area in and around the MaT with proper techniques it is now time to get the adhesive mixed. The adhesive kit consists of 2 separate parts that must be mixed in the correct ratio. The ratio is given by the vendor (see Table 6). To mix it properly the adhesive has to be mixed by weight. Figure 55 shows the three steps of mixing the adhesives. In the first two steps the weight has to be measured. Part A is determining the added part B. For example 100g of part A would need a mix with 22g of part B. The last step is to mix the two parts together to get a purple colored paste.

Abbildung in dieser Leseprobe nicht enthalten

Figure 55: Scale is used for mixing the Hysol9309.3NA

The mixed adhesive has a consistence that is comparable with smooth honey. It is easy to process and spread on the desired surface. Therefore the manufacturer uses a brush to spread the adhesive on the bonding area. One test coupon will be brushed with 3-4g of adhesive to get all the desired surfaces bonded together. Compared to the first test this is already approximately 8x less weight. Figure 56 is demonstrating how the adhesive is applied on the surface. Thereby it is important to apply the Hysol9309.3NA in width of the mortise. Otherwise the side edges of the panel with the tenon will not bond correctly to skin and lose in strength.

Abbildung in dieser Leseprobe nicht enthalten

Figure 56: Applying the adhesive

The MaT panels can now be joined together and hold in position. To hold them in a perpendicular position the manufacturer has to install screws from the back of the panel into the panel with the tenon. The process is illustrated in Figure 57. The filled mortise is also visible through the face sheet. Therefore it is always a control tool to ensure a correct manufacturing.

Abbildung in dieser Leseprobe nicht enthalten

Figure 57: Screws ensure a perpendicular position

The completed test coupons shall be stored in the heating room at 180° Fahrenheit for a 6 hour cure or left in the hangar at room temperature (72° Fahrenheit) to cure for 3-5 days.

4.4.3 Evaluation of the test results

The third test is the result of month-long development process and research. After evaluating and analysing the first and second tests the joint is getting more and more efficient. The results are exceeding the expectations and demands that are given at the beginning of this project. Although the last test showed problems in the longitudinal directions because the expected strength was lower than the actual strength, the tension test ran perfectly. This chapter gathers all the information of the results from the last test and analysis the problems and the valuable properties. For the last test it is only designated to test T joints. The L joints will have similar results to the T joints with the new technique to bond the side edges. With depleted company tests and limited budget it is important to get the remaining load directions tested. Therefore we will test three test coupons for each load directions that will be manufactured in the same way. This chapter will be divided into the three load directions to analyse each separately.

4.4.3.1 Tension test

The first load direction that will be tested is tension. The main load direction for this project has priority and was the crucial factor in the previous tests. The newest test coupon should be able to withstand loads that are higher than any load in the previous tests. The test setup and the method will not change with the next tests.

Abbildung in dieser Leseprobe nicht enthalten

Figure 58: The first test sample (weakest)

that account only 1.3% of the mean value (the comparable panel pin joint had 24.5%[37]). This low deviation will bring high allowables that are useful especially with a high A-basis allowable. The calculated allowables for this test are 773lbf for the B-basis and 585 for A-basis in regard to the Weibull distribution. This A-basis is 5 times higher than the allowable for the similar panel joint with panel pins. Although this calculation is only made with 3 test coupons and not with 12 the results are promising. As described in Chapter 4.2.2.1 the allowables of the panel pins and the MaT are not really comparable because the allowables of three coupons are lower than they could be with more tested coupons with the same consistency.

Figure 58 shows the first sample and its shape after the test. Several positive perceptions can be made with this destroyed test coupon for example that no broken adhesive is actuator for the failure. This happened on most of the other test specimen of the first two tests. Even though not all of them were resulted in a maximum failure load it is part of the failure. Another positive result is the failed glass fiber where the side edges were bonded to the panel. The bonded core is enhancing the maximum strength as it is illustrated in Figure 59. Especially the left side has ripped out the core min. to the length of 1 hexagon into the panel. This test specimen is the third and the strongest yet.

Abbildung in dieser Leseprobe nicht enthalten

Figure 59: The third test sample (strongest)

The consistency can be shown with the diagram of the test coupon -016 (see Figure 60). Because the third test coupon has a higher maximum load the deflection is deviated from the other two test coupons. Therefore only the first and second test coupon will be analyzed in the following. However the failure modes are the same.

0-0.19 inches: The load is constantly increasing with growing deflection. The distribution on both test specimens is very similar. The first small drops are internal minor failures.

0.19-0.2 inches: With reaching the maximum strength both test specimen are failing due to skin failure on the side edges at loadmax§ OEI 7KH ORDG GHFUHDVHV LQVWDQWO\ DQG recovers at a rate of approximately 420 lbf. This recover is caused by the MaT and the other side edge that still are capable of holding a load that is on the level of maximum panel pin strength.

0.2-0.3 inches: The load is increasing again after the first failure and is now pulling the MaT apart. To accomplish this, the skin has to crack around the mortise. With a Hysol9309.3NA MaT joint this is approximately 600 lbf.

0.3-0.36 inches: After reaching the second failure load of 600 lbf the second side edge fails and rips the MaT apart. The following loads are not considered.

Abbildung in dieser Leseprobe nicht enthalten

Figure 60: Data of -016 in a diagram

All in all the test results are more than satisfying. With an easy and fast method of manufacturing, low weight and strong allowables this MaT joint is meeting the FAA regulations and can be a promising option for the panel pins.

4.4.3.2 Longitudinal Shear

Before the MaT joint reaches the next stage of development which includes the test plan and detailed process specifications the testing for the other two load directions have to be performed and analyzed. The longitudinal shear is tested on the same test setup as the tension or the transverse shear. Whereas the panel with the mortise will only be clamped on the tension and transverse shear samples, the longitudinal shear samples are controlled with pins on both panels (see Figure 61). The left picture shows the panel with the tenon. The pins that are inducted into the panel will pull in the direction of the holes in order to get a longitudinal shear force.

Abbildung in dieser Leseprobe nicht enthalten

Figure 61: Test setup for longitudinal shear

The testing of the first test coupons had a sobering result. The joint was stronger than expected and the failure was not in the adhesive but in the holes of the test coupon. Figure 62 visualizes the failure mode of the longitudinal shear test coupons. The withstood load in the MaT joint was so high that the bearing stress in the holes exceeded the limit.

Abbildung in dieser Leseprobe nicht enthalten

Figure 62: Unexpected failure of -018-003

The maximum loads can be seen in Table 20 and are compared to the panel pin failure load already more than twice as high. Since the test coupons were not tested to the desired failure mode they have been modified to withstand the bearing stress.

Abbildung in dieser Leseprobe nicht enthalten

Table 20: Failed test results of the longitudinal shear test specimen

Two metal sheets are mounted to the test coupons from either side to ensure enough strength to fail on the desired point of failure which would be the bond line. With that, the modified test coupons failed successfully and can therefore be properly analyzed. Different to most of the tests that are done the results of this test have a high standard deviation because of the third test run that has a 200lbf lower maximum strength. Table 21 is showing the properties of the three longitudinal test coupons and the results of the test. Although the standard deviation is very high with these test coupons they still are more than twice as high and the allowables are much better. The biggest advantage is the fact that the panel pins have no A-basis allowable[38] for longitudinal shear because the value is negative. Therefore the stress engineers always have to proof a second safe load distribution. This additional effort could be avoided by using MaT.

Abbildung in dieser Leseprobe nicht enthalten

Table 21: successful test results of the longitudinal shear test specimen

The Figure 63 illustrates the failures of the strongest and the weakest test specimen for longitudinal shear. The various appearances between the two samples is because the

Abbildung in dieser Leseprobe nicht enthalten

Figure 63: Failed sample -018 (Top is the weakest and bottom the strongest joint)

load input on the stronger sample stopped right after the test coupon broke. This is also visible in the diagram which is described below. What made the two test coupons deviated in a range of 200 lbf is the incomplete bonding of the side edges. The skin of the weaker test coupon is not fully attached to the panel with adhesive. This leads to a weaker joint and therefor it can be said that this test coupon is manufactured incorrectly.

The diagram in Figure 64 is giving details about the behavior of the samples during the test. The first thing that catches the eye is the shifted curve of the first test run. Even though it broke in the area of the expected load the deflection at the first failure was about 0.15 inches higher than the other samples. Additionally the curves are smooth and nearly linear in the phase of increasing load and in the phase of failure.

The following analysis of the behavior at certain deflections is only referring to the second and third test run because the shift of the third test run is too high.

0-0.25 inches: The load is constantly increasing with growing deflection. The distribution on both test specimens is very similar. The first small drops are internal minor failures.

0.25-0.28 inches: The first failure appears due to reaching the maximum shear strength of the joint. The high value can be led back to the shear force that occurs in the direction of the bond line.

0.28-0.8 inches: The behaviour of this test is unique in this project because the failure results in a steep drop to a minimum shear load. Nevertheless the third test has a small increase of shear strength because the MaT are blocking each other. This small increase has no impact on the maximum shear strength that is caused by the adhesive bond line on the side edges.

Abbildung in dieser Leseprobe nicht enthalten

Figure 64: data of -018 in a diagram

All in all the longitudinal shear test was successful due to the high maximum value. The inconsistency can compensate by that to achieve still valuable allowables. The big advantage with MaT compared with the panel pins is that the it has an A-basis allowable that can be used for stress analysis in designed cabinets.

4.4.3.3 Transverse Shear

The transverse shear is not expected to be very strong because the load is not following the path of the bond line. Since it is perpendicular to the directions of the side edges the adhesive cannot get full resistance to shear loads. Nevertheless the tenon will block inside the mortise and all together could cause a sufficient result. Recording to the stress engineers this load direction is the least critical in the stress analysis of a cabinet. The high shear strength of Hysol is important for this test. The test results can be inspected in

Table 22.

Abbildung in dieser Leseprobe nicht enthalten

Table 22: Test results of -017

The results are throughout consistent which is positive. Although the maximum shear strength is not as high as the panel pins it sure has a better standard deviation. Whereas the panel pins had a standard deviation of 54.6[39] the MaT test coupons only have 4.6. In a comparison the allowables of the MaT are as good as panel pins. The tested samples are shown in Figure 65 with the stronger sample on the top and the weaker sample on the bottom. With the process of pulling on the panel with the pins the tenon got ripped off. The adhesive inside the mortise was strong enough to withstand the tension that is forced in the face sheet with increasing deflection. The white stripes on the skin of the panel are a characteristic of a skin failure due to bending. The lower side of the panel has a skin that is intact. This indicates that only the upper panel had very strong bending stress. This fact raises the question whether the adhesive has failed only due to shear loads. With a skin that failed first it could be that the adhesive got an additional peeling load. The peel strength of an adhesive is very low and would influence the maximum strength dramatically.

Closer analysis with the video material and the diagram shown in Figure 66 has revealed that the bending of the panel and the associated peeling happened after the maximum shear strength has been reached. Therefore it can be said that the test coupon exclusively had to withstand shear loads. In this test the core had the most impact because it extended the area in direction to the shear load.

Abbildung in dieser Leseprobe nicht enthalten

Figure 65: tested transverse shear samples

This test showed the importance of the deflection-load diagram for analysis of tests with different failures. Because is needed to separate the panels the failure modes can occur even after the maximum strength has been reached. To proof the statements the diagram in Figure 66 will be analysed with increasing deflection.

Abbildung in dieser Leseprobe nicht enthalten

Figure 66: test data of -019 in a diagram

Because all the three test coupons do not differ much in their characteristic the following analysis is related to all three

0-0.3 inches: In contrast to the longitudinal shear the transverse shear has no linear character in the first few inches of deflection. Since this joint is not very stable for transverse shear many small adhesive failures occur before even the maximum strength has been reached.

0.27-0.32 inches: The first big area of adhesive failed and is setting the maximum of the shear load. The load is only slowly decreasing after the first failure which is uncommon.

0.32-0.7 inches: The skin might broke for the first failure but the core is still holding to cause a small decrease of load that is applied to the sample. The scattered curve is caused by small adhesive bondings that are breaking. The increasing strength at 0.6 inches at the third test run is the tenon that begins to be pulled in tension.

0.7-0.78 inches: The bending of the panel has reached the maximum and the MaT is holding the load until the skin tears and gets the load to drop to a minimum.

The transverse shear test has shown the weaknesses of the bonded joint. The advantage in this load direction is zero but it is not a disadvantage. With the transverse shear load direction as one of the not critical parts for designing a cabinet it is safe to say that the MaT has shown the capability of being stronger than the panel pins. If a problem accurse with an insufficient area of a cabinet in regard to transverse shear it is possible to place a panel pin in the critical area. The FAA approval for panel pins is already so it is always possible to enhance the MaT joint in weak spots.

4.5 Bending test

4.5.1 Introduction

For proving that the MaT has a minor effect on the bending strength for panels in a cabinet, two test coupons were designed to show compliance. The four point bending test is a common method to calculate the bending strength of a material. The bending test specimen is placed on two supports and will be bent with two forces of the same amount in the same distance (a) from the support. The theoretical setup is shown in

Figure 67.

Abbildung in dieser Leseprobe nicht enthalten

Figure 67: 4 point bending setup (theoretical)

The actual test setup is shown in Figure 68 and contains the untreated panel on the left picture. It is a 3/4 inch Teklam panel with 0.02 inch thick face sheets. The right picture illustrates the same panel with a MaT in it. It is bonded with Hysol9309.3NA in the mortise and in the side edges to use the technique that this project has resulted in. The panel has two different load types in the four point bending test. The upper side of the panel will have a compression load whereas the lower side will have a tension load. Which side is weaker will be evaluated on the plain test specimen. Although it is assumed to be the compression side the plain test specimen will show the result anyway.

Abbildung in dieser Leseprobe nicht enthalten

Figure 68: plain sample (left) and MaT sample (right) in the setup

4.5.2 Test results

As described in the previous chapter the plain test specimen will be tested first. The failure mode for the plain test was as expected due to compression on the upper side. Figure 69 shows the sample right after the failure. The right side of the right load has cracked on the upper face sheet. The inner side of the left load has buckled the face sheet off of the core. The maximum failure load for this test was 1510 lbf.

Abbildung in dieser Leseprobe nicht enthalten

Figure 69: Failure of the plain panel

Subsequently the bending test with the installed MaT will be tested. The first test has resulted that the compression side is the weaker side. For this reason the MaT will be tested with the cut mortise on the upper side. Since the MaT could be installed on the compression or the tension side in reality it has to be ensured to test the weaker joint.

The failure of the MaT bending test can be seen in Figure 70 and differs from the first test to the effect that the panel breaks at the MaT. This point of failure is anything but a surprise because the lower width of the panel made it weaker.

Abbildung in dieser Leseprobe nicht enthalten

Figure 70: Failure of the MaT panel

The maximum load of the MaT bending panel was 1200 lbf. To compare both panels in its strength the face sheet strength now needs to be calculated. This is the critical allowable for designing a honeycomb panel with bending. The product data sheet for Teklam panels is calling out the allowables for the facing stress in a long beam test[40].

Abbildung in dieser Leseprobe nicht enthalten

Figure 71: Free body diagram for the bending test

The equation for the face sheet strength is:

Abbildung in dieser Leseprobe nicht enthalten

To calculate the bending moment for each test we need to know the dimensions of the test setup and the free body diagram seen in Figure 71. The changing variable for both tests is the maximum force that is applied on the panel. The dimensions are designed to be as simple as possible with a total length of ܮ = 20 ݄݅݊ܿ݁ݏ. The maximum bending moment for the plain test coupon can be evaluated with the following equation:

Abbildung in dieser Leseprobe nicht enthalten

The maximum bending moment for the MaT test coupon would be then:

Abbildung in dieser Leseprobe nicht enthalten

After calculating the maximum bending moments for both tests it is now to figure out the face sheet stress that occur in both tests. The face sheet stress can be evaluated with:

Abbildung in dieser Leseprobe nicht enthalten

Where b is the width of the panel, tf is the thickness of the face sheet and h is the distance between the center of each face sheet (see Figure 72)

Abbildung in dieser Leseprobe nicht enthalten

Figure 72: distance between face sheets

To calculate the facing stress of each test coupon we now have to apply the right values into the equation:

Abbildung in dieser Leseprobe nicht enthalten

Although the maximum moment of the MaT joint is smaller the facing stress is increasing due to the lower width of the panel. With the mortise being 1.5 inches in width the total width of the panel is now only 6.5 inches. This results in the following equation:

Abbildung in dieser Leseprobe nicht enthalten

To get the total facing stress the calculated value has to be multiplied by two since the honeycomb consists of two face sheets.

Abbildung in dieser Leseprobe nicht enthalten

The MaT lowers the percentage of facing stress in the face sheet by approximately 1%. This shows that the mortise inside the panel with bonded tenon has a minor impact on the bending strength of the panel. This test showed that for future designs with MaT in cabinets would equal the current designs with panel pins in regard to bending strength.

5 Conclusion

5.1 Results

This paper is a design analysis for a new panel joint that will be used for cabinets and monuments in the VIP completion. It proves that using panel pins for joining two panels is not the strongest and most effective technique. With substantial progress in the adhesive research and modern CNC machinery it was about to combine different variables to achieve a mortise and tenon joint that not only meets the requirements of the Federal Aviation Administration but also achieves the goal of this project which was to be more effective than the panel pins in different categories. The first category was to lower the cost of the joint and it has been achieved to be 15% cheaper per joint. This value only includes the fixed costs per joint and not includes the major fact that a decent amount of work hours will be saved. This is being proven by the minimization of time exposure in the manufacturing process. The CNC is cutting the mortise and tenon to fit perfectly and additional process specification will lead to less human errors. The steps and tools needed for the manufacturing process have been decreased to ensure a fast and secure production of cabinets and monuments. The third major improvement and requirement to end this project successfully was to lower the weight of one joint and therefore the weight of the whole monument. With the MaT joint can be saved up to 75% of the weight compared with the panel pins. Especially in the aviation industry the weight has a high significance and needs to be lowered as much as possible to minimize the operating costs for the holder of the aircraft. The last and most important fact was the improving of strength in the joint. To accomplish the requirement this paper is describing a research and developing process that is using certain procedures to analyse the performed tests. The used methods led to an increase of the strength in tension and longitudinal shear. The mean value of the tension tests were 75% stronger than the panel pins and had a much higher consistency. This would lead to an A-basis allowable that is probably four times higher due to the lower standard deviation. The strength of the longitudinal shear test coupons had a 160% higher value but a consistency comparable to the panel pins. Nevertheless this led to a positive A-basis allowable (panel pin is negative) which saves a work hours on the stress engineers. The transverse shear allowable was weaker in strength but better in consistency which can be seen as compensation.

All in all the project was a success and generated a new joint that has better values in weight, strength, time exposure, manufacturability and cost.

5.2 Outlook

After the first tests for designing a new panel joint were made it is still a long way to have the MaT joint ready for the integration in aircrafts. The focus in this test was on T joints and the related tension strength whereas the improvements on the L joints still have to be made. Although the major research and development has been made and solutions have been presented the new L joints still have to be tested to show the predicted strength. After that the test plan with the appropriate process specification has to be designed in order to get FAA approval. For each joint combination must be at least twelve specimen tested.

The engineering and testing for this joint were made and the goal of the project has been reached. The future will show whether the new technique for connecting honeycomb panels will find its way into the VIP completion.

List of References

Abbildung in dieser Leseprobe nicht enthalten

Attachment A

Abbildung in dieser Leseprobe nicht enthalten

Attachment B

§ 25.603 Materials.

The suitability and durability of materials used for parts, the failure of which could adversely affect safety, must—

(a) Be established on the basis of experience or tests;

(b) Conform to approved specifications (such as industry or military specifications, or Technical Standard Orders) that ensure their having the strength and other properties assumed in the design data; and

(c) Take into account the effects of environmental conditions, such as temperature and humidity, expected in service.

[Doc. No. 5066, 29 FR 18291, Dec. 24, 1964, as amended by Amdt. 25-38, 41 FR 55466, Dec. 20 1976; Amdt. 25-46, 43 FR 50595, Oct. 30, 1978]

§ 25.605 Fabrication methods.

(a) The methods of fabrication used must produce a consistently sound structure. If a fabrication process (such as gluing, spot welding, or heat treating) requires close control to reach this objective, the process must be performed under an approved process specification.

(b) Each new aircraft fabrication method must be substantiated by a test program.

[Doc. No. 5066, 29 FR 18291, Dec. 24, 1964, as amended by Amdt. 25-46, 43 FR 50595, Oct. 30, 1978]

§ 25.613 Material strength properties and design values.

(a) Material strength properties must be based on enough tests of material meeting approved specifications to establish design values on a statistical basis.

(b) Design values must be chosen to minimize the probability of structural failures due to material variability. Except as provided in paragraph

(e) of this section, compliance with this paragraph must be shown by selecting design values which assure material strength with the following probability:

(1) Where applied loads are eventually distributed through a single member within an assembly, the failure of which would result in loss of structural integrity of the component, 99 percent probability with 95 percent confidence.

(2) For redundant structure, in which the failure of individual elements would result in applied loads being safely distributed to other load carrying members, 90 percent probability with 95 percent confidence.

(c) The effects of temperature on allowable stresses used for design in an essential component or structure must be considered where thermal effects are significant under normal operating conditions.

(d) The strength, detail design, and fabrication of the structure must minimize the probability of disastrous fatigue failure, particularly at points of stress concentration.

Attachment B 82

(e) Greater design values may be used if a “premium selection” of the material is made in which a specimen of each individual item is tested before use to determine that the actual strength properties of that particular item will equal or exceed those used in design.

[Doc. No. 5066, 29 FR 18291, Dec. 24, 1964, as amended by Amdt. 25-46, 43 FR 50595, Oct. 30, 1978; Amdt. 25-72, 55 FR 29776, July 20, 1990]

§ 25.853 Compartment interiors.

For each compartment occupied by the crew or passengers, the following apply:

(a) Materials (including finishes or decorative surfaces applied to the materials) must meet the applicable test criteria prescribed in part I of appendix F of this part, or other approved equivalent methods, regardless of the passenger capacity of the airplane.

(b)[Reserved]

(c) In addition to meeting the requirements of paragraph (a) of this section, seat cushions, except those on flight crewmember seats, must meet the test requirements of part II of appendix F of this part, or other equivalent methods, regardless of the passenger capacity of the airplane.

(d) Except as provided in paragraph (e) of this section, the following interior components of airplanes with passenger capacities of 20 or more must also meet the test requirements of parts IV and V of appendix F of this part, or other approved equivalent method, in addition to the flammability requirements prescribed in paragraph (a) of this section:

(1) Interior ceiling and wall panels, other than lighting lenses and windows;

(2) Partitions, other than transparent panels needed to enhance cabin safety;

(3) Galley structure, including exposed surfaces of stowed carts and standard containers and the cavity walls that are exposed when a full complement of such carts or containers is not carried; and

(4) Large cabinets and cabin stowage compartments, other than underseat stowage compartments for stowing small items such as magazines and maps.

(e) The interiors of compartments, such as pilot compartments, galleys, lavatories, crew rest quarters, cabinets and stowage compartments, need not meet the standards of paragraph (d) of this section, provided the interiors of such compartments are isolated from the main passenger cabin by doors or equivalent means that would normally be closed during an emergency landing condition.

(f) Smoking is not to be allowed in lavatories. If smoking is to be allowed in any other compartment occupied by the crew or passengers, an adequate number of self-contained, removable ashtrays must be provided for all seated occupants.

(g) Regardless of whether smoking is allowed in any other part of the airplane, lavatories must have self-contained, removable ashtrays located conspicuously on or near the entry side of each lavatory door, except that one ashtray may serve more than one lavatory door if the ashtray can be seen readily from the cabin side of each lavatory served.

(h) Each receptacle used for the disposal of flammable waste material must be fully enclosed, constructed of at least fire resistant materials, and must contain fires likely to occur in it under normal use. The capability of the receptacle to contain those fires under all probable conditions of wear, misalignment, and ventilation expected in service must be demonstrated by test.

§ 25.561 Emergency Landing conditions

(a) The airplane, although it may be damaged in emergency landing conditions on land or water, must be designed as prescribed in this section to protect each occupant under those conditions.

(b) The structure must be designed to give each occupant every reasonable chance of escaping serious injury in a minor crash landing when—

(1) Proper use is made of seats, belts, and all other safety design provisions;
(2) The wheels are retracted (where applicable); and
(3) The occupant experiences the following ultimate inertia forces acting separately relative to the surrounding structure:

(i) Upward, 3.0g
(ii) Forward, 9.0g
(iii) Sideward, 3.0g on the airframe; and 4.0g on the seats and their attachments.
(iv) Downward, 6.0g

(v) Rearward, 1.5g

(c) For equipment, cargo in the passenger compartments and any other large masses, the following apply:

(1) Except as provided in paragraph (c)(2) of this section, these items must be positioned so that if they break loose they will be unlikely to:

(i) Cause direct injury to occupants;
(ii) Penetrate fuel tanks or lines or cause fire or explosion hazard by damage to adjacent systems; or
(iii) Nullify any of the escape facilities provided for use after an emergency landing.

(2) When such positioning is not practical (e.g. fuselage mounted engines or auxiliary power units) each such item of mass shall be restrained under all loads up to those specified in paragraph (b) (3) of this section. The local attachments for these items should be designed to withstand 1.33 times the specified loads if these items are subject to severe wear and tear through frequent removal (e.g. quick change interior items).

(d) Seats and items of mass (and their supporting structure) must not deform under any loads up to those specified in paragraph (b)(3) of this section in any manner that would impede subsequent rapid evacuation of occupants.

[Doc. No. 5066, 29 FR 18291, Dec. 24, 1964, as amended by Amdt. 25-23, 35 FR 5673, Apr. 8, 1970; Amdt. 25-64, 53 FR 17646, May 17, 1988; Amdt. 25-91, 62 FR 40706, July 29, 1997

Appendix F to Part 25

Part I—Test Criteria and Procedures for Showing Compliance with §25.853, or §25.855.

(a) Material test criteria —(1) Interior compartments occupied by crew or passengers. (i) Interior ceiling panels, interior wall panels, partitions, galley structure, large cabinet walls, structural flooring, and materials used in the construction of stowage compartments (other than underseat stowage compartments and compartments for stowing small items such as magazines and maps) must be self-extinguishing when tested vertically in accordance with the applicable portions of part I of this appendix. The average burn length may not exceed 6 inches and the average flame time after removal of the flame source may not exceed 15 seconds. Drippings from the test specimen may not continue to flame for more than an average of 3 seconds after falling.

(ii) Floor covering, textiles (including draperies and upholstery), seat cushions, padding, decorative and nondecorative coated fabrics, leather, trays and galley furnishings, electrical conduit, air ducting, joint and edge covering, liners of Class B and E cargo or baggage compartments, floor panels of Class B, C, D, or E cargo or baggage compartments, cargo covers and transparencies, molded and thermoformed parts, air ducting joints, and trim strips (decorative and chafing), that are constructed of materials not covered in subparagraph (iv) below, must be self-extinguishing when tested vertically in accordance with the applicable portions of part I of this appendix or other approved equivalent means. The average burn length may not exceed 8 inches, and the average flame time after removal of the flame source may not exceed 15 seconds. Drippings from the test specimen may not continue to flame for more than an average of

5 seconds after falling.

(iii) Motion picture film must be safety film meeting the Standard Specifications for Safety Photographic Film PHI.25 (available from the American National Standards Institute, 1430 Broadway, New York, NY 10018). If the film travels through ducts, the ducts must meet the requirements of subparagraph (ii) of this paragraph.

(iv) Clear plastic windows and signs, parts constructed in whole or in part of elastomeric materials, edge lighted instrument assemblies consisting of two or more instruments in a common housing, seat belts, shoulder harnesses, and cargo and baggage tiedown equipment, including containers, bins, pallets, etc., used in passenger or crew compartments, may not have an average burn rate greater than 2.5 inches per minute when tested horizontally in accordance with the applicable portions of this appendix.

(v) Except for small parts (such as knobs, handles, rollers, fasteners, clips, grommets, rub strips, pulleys, and small electrical parts) that would not contribute significantly to the propagation of a fire and for electrical wire and cable insulation, materials in items not specified in paragraphs

(a)(1)(i), (ii), (iii), or (iv) of part I of this appendix may not have a burn rate greater than 4.0 inches per minute when tested horizontally in accordance with the applicable portions of this appendix.

Attachment C

Abbildung in dieser Leseprobe nicht enthalten

Attachment D

Abbildung in dieser Leseprobe nicht enthalten

AEROK 9110R

PRODUCT

Aerok 9110R is a two component low density adhesive/filler developed for bonding composite and aluminium honeycomb panels.

TECHNICAL CHARACTERISTICS

Abbildung in dieser Leseprobe nicht enthalten

(*) These values may vary depending on environmental factors such as: temperature, moisture and type of substrates. Values will also vary due to volume of material mixed/used due to exothermic reaction

Attachment E

Abbildung in dieser Leseprobe nicht enthalten

Attachment F

Abbildung in dieser Leseprobe nicht enthalten

Attachment G

Abbildung in dieser Leseprobe nicht enthalten

Attachment H

Acceptable Methods of Compliance

Abbildung in dieser Leseprobe nicht enthalten

Attachment I

Abbildung in dieser Leseprobe nicht enthalten

Attachment J

Abbildung in dieser Leseprobe nicht enthalten

---

[1] FAR1

[2] IRR 2013

[3] FAR 5

[4] FAR 1

[5] FAR 6

[6] PS-202

[7] FAR 2

[8] FAR 3

[9] FAR 4

[10] BizJet 2010-1

[11] Norbond™

[12] Bizjet 2010-2

[13] Carbide Depot

[14] Aluminum

[15] Shear allowable

[16] MSC

[17] Bending test

[18] Norbond™

[19] AEROK

[20] FAR 6

[21] HYSOL 1

[22] Tension

[23] Compressive

[24] Lap shear

[25] The costs are evaluated of the purchased history of BizJet and converted into liters

[26] ATR

[27] HYSOL

[28] AEROK

[29] The vendor has stated that they currently have no tension allowable

[30] Tolerances

[31] Standard deviation

[32] B-basis

[33] CMH

[34] FAR7

[35] FAR8

[36] HYSOL 2

[37] Aeronautique-1

[38] Aeronautique-2

[39] Aeronautique-3

[40] TEKLAM