General Discussion

1.  Thermal-Cycle Endurance of Bonded Joints

Posted 03-12-2018 15:34

Matthew Nahan
Lockheed Martin Space Systems Co
Decorah IA

2.  RE: Thermal-Cycle Endurance of Bonded Joints

Posted 03-13-2018 12:30
Edited by Sandesh Hegde 03-17-2018 10:42
Hello Matthew,
The work we do is more focussed on the effect of thermal fatigue on the mechanical properties of composite honeycomb sandwich structures, with CFRP facesheet and Kevlar core.
In our case, we observe the formation of microcrack, on the core-facesheet interface region of the sandwich structures, due to the difference in coefficient of thermal expansion (CTE), of the constituents used in the sandwich material. We observed that the microcracks don't grow much after 40 thermal cycles based on laboratory experimental observations and as a result, the flatwise tensile strength also doesn't reduce much after 40 thermal cycles.
You may consider the above observations for bonding using adhesives, am not sure if it is relevant for metal Fasteners for bonding composites.
Good day!
Kind Regards,

Sandesh Hegde

MSc student, Concordia Centre for Composites,

Department of Mechanical, Industrial and Aerospace engineering,

Concordia University, Montreal, Canada.

Phone Number: +1 514 5622 749


3.  RE: Thermal-Cycle Endurance of Bonded Joints

Posted 03-16-2018 13:06

Although I expect thermal cycling effects between face sheet and core to be an issue separate from thermal cycling effects upon bonded joints; I am glad you brought-up the subject.  I am also experiencing such a sandwich problem.  We know that thermal loading produces stress concentrations, at free edges and corners between face sheet and core.  And we know from test that thermal-cycling can dramatically effect residual strength as measured using the flat-wise tension test (FWT).  We're taking a two-prong approach to achieve higher strength: 1) improve the bonding process, and 2) switch from to a tougher face sheet ply material system.  My process is too much trial-and-error for my liking.

We have some general questions too: 1) is the FWT test the best method?, 2) if a panel survives 40 cycles, will it survive 2000? 3) do tough and compliant buffer plies help when placed between core and face sheet?, 4) is there any benefit to adding more adhesive or increasing the resin richness of the first ply?

I see that you have a SAMPE paper coming out on the issue and I will look forward to reading it.

Matthew Nahan
Sr. Mechanical Engineer
Lockheed Martin Space Systems Co
Decorah IA

4.  RE: Thermal-Cycle Endurance of Bonded Joints

Posted 30 days ago
Matthew, Sandesh,

There are a few different useful points I can add to your discussion.  I'd be happy to discuss things with you more, of course.

1.  There are not always serious fatigue concerns in space structures (e.g., composite bonded joints) because the number of cycles is typically low for a static structure.  Note that I am not talking about actuators, motors, cryo-coolers, etc.
2.  Once you've survived one cycle, I think you would be more concerned with microcracking and any loss of strength, dimensional stability, or stiffness.  For my work with very cold structures I thought this was helpful:  "Timmerman J F, Matthew S T. Matrix and fiber influence on the cryogenic microcracking of carbon fiber/epoxy composites[J]. Composites: Part A, 2002, 33(3): 323-329"  One key finding here was that a "handful" of cycles was enough (and often we are required to have at least 7 or 8 cycles on spaceflight hardware)
3.  I'm surprised you are able to use a partially debonded item, though perhaps that disbond did not occur in development.  Perhaps you are talking about flight articles that already exist.
4.  I'm also giving a paper at SAMPE and it centers around recovering from a debond between facesheet and core and how to improve the strength during an extreme thermal cycle down to -396 F (-238C).  I hope you can make it, it's called "

Development of a Solar Array substrate for the Europa environment".
5.  I do believe FWT is your best test method for an issue between facesheet and core.
6.  Toughness of your film adhesive can make a major difference.
7.  Consider your surface preparation and depth of abrasion.  We found less resin to be better than more resin (we abraded until some "black" was present, from the carbon fiber).
8.  Recall that composites are not homogeneous and are composites of two different materials, each with unique allowables.  A debond would typically be film-adhesive or matrix-driven.  You should work with failure criteria which can handle matrix-driven failures (We adapted SIFT, or Onset Theory, to our application and I am very much a believer in this criteria for this kind of situation).  This will inform you about the temperature limits of your system and will show you just how much your design window shrinks if you are working at extreme temperature limits.

John Troll
Senior Mechanical Engineer
Genesis Engineering Solutions, Inc.
Lanham, MD

5.  RE: Thermal-Cycle Endurance of Bonded Joints

Posted 27 days ago

Thanks for your generous and thoughtful response.

We are not planning to fly anything with the disbonds we're experiencing.

In my current bonded joint application, our joints cracked but did not technically fail (i.e. still carried required load).  After less than 40 thermal cycles, we lost maybe 25% of our bond area.  Given a lack of testing of composite bonded joints under extended thermal-cycle conditions, I lack your confidence that this is not a low-cycle fatigue issue.  I agree that microcracking is a significant factor at play, maybe for the problem of core-to-face sheet disband as well.

I eagerly await your paper.

Matthew Nahan
Sr. Mechanical Engineer
Lockheed Martin Space Systems Co
Decorah IA

6.  RE: Thermal-Cycle Endurance of Bonded Joints

Posted 19 days ago
Matthew, John

Thank you very much for the information. To be honest, so far we had been working only on the causes for microcrack formation and its effect on the mechanical strength, I am glad you brought up various approaches to address thermal cycling-related issues, in the sandwich structures.

Some information regarding Flatwise tensile test are below
1) Yes, as John pointed out, this is the best test to address issues between faceseet and core.
2) The adhesive we used, to bond the test specimen to the loading block of the test fixture is LOCTITE EA 9392 AERO.
3)  Before applying adhesive the sample's facesheet surface needs to be lightly sanded (we used 320 grit size sandpaper).
4)  Make sure the sandwich coupons have laminates that are properly cut/damage free/laminates properly aligned top to bottom
5) Test setup as per ASTM C297 is key; proper time/rate of speed, Coupon size (with respect to cell size etc) and the test setup with orthogonal pins will eliminate the most variation in the results.

Good Day!

Sandesh Hegde

7.  RE: Thermal-Cycle Endurance of Bonded Joints

Posted 16 days ago

Two things that I have found beneficial to life cycle testing include conditioning environment and scale of sample/specimens.

With regard to environment, it is not possible to recreate the exact conditions that a structure will experience.  However, exposing samples to different extremes or boundaries of the intended application will highlight sensitivities and hint at strategies to improve.  For example, condition samples hot/wet, hot/dry in air, hot/dry in vacuum, hot/dry in rarified gas.  Measure sample weight as a function of time.  Condition both cored laminates as well as the skins, adhesive, and core individually. Cross section the samples periodically to inspect the progression of degradation in from the surfaces and edges.

Epoxies are quite stable at elevated temperatures under vacuum, but as the concentration of moisture, oxygen, and or other reactive materials increases, the rate of mass loss also increases dramatically.  The benefits of even a very thin sacrificial coating are impressive.

Often degradation is primarily due to environmental conditions which in turn are a function of diffusion rates, permeability, and surface area.  Samples need to be several times bigger than the test specimens to control and characterize edge effects.  For my research, at least 2 inches of sample length (distance to the nearest edge) for every week (168hrs) at 350F (177C) to observe the "wavefront" of oxidation.

I look forward to reviewing the papers.



Jed Brich
Research Engineer
Janicki Industries

8.  RE: Thermal-Cycle Endurance of Bonded Joints

Posted 15 days ago

An aromatic or aliphatic zirconate is an excellent adhesion promotor for graphite – see link: .


Salvatore J. Monte

P.E., B.C.E., M.S.-Polymeric Materials, Fellow (SPE)


Kenrich Petrochemicals, Inc.

PO Box 32

Bayonne, NJ 07002 USA

Tel No. 201-823-9000

Fax No. 201-823-0691



Customer Support:


CONFIDENTIALITY NOTICE. This email message and any attachments may be confidential and privileged. If you are not the intended recipient, please do not read, copy or re-send this email message or its attachments to any third party recipient; immediately notify the sender by reply email or by collect call to 201.823.9000; and delete this email message and any attachments. Thank you for your assistance.


9.  RE: Thermal-Cycle Endurance of Bonded Joints

Posted 14 days ago
If anyone is ever interested in understanding adhesion properties which would relate to our discussion, I encourage you to check out my company's website at Advancing your Surface Science - KRÜSS GmbH to learn about the various instruments we offer that provides information for concepts like work of adhesion, polar bonding, bond strength.  If you are ever interested in discussing more, I am happy to assist.

Mark McCarthy
KRÜSS Scientific Instruments, Inc.
Matthews NC