ILSS Failures - Processing Factors For Low Test Scores

[Obsidian]

Evening Forum ,

Can anyone suggest processing factors that may yield lower than expected ILSS test scores for CFRP laminates . The laminates are usually aproximately 8 plies thick with a nominal thickness of 2mm . The fabric is usually woven with the 0 Deg direction aligned in the same orientation for all 8 plies . The resin is an epoxy . The material is laid up as a pre-impregante and autoclave cured at 360 F and 90 psi . The Vf is typically 60 % and Vv is < 2% .

Any thoughts would be appreciated .

Cheers

 

[RPstress]

Mostly obvious stuff, I'm afraid:

What was the NDT like?

Has the processing data been checked (thermocouples, pressure and vacuum)?

Is this a multidirectional laminate or is it all 0 degrees (and what sort of laminates are the expected strengths for)?
(See thread327-145748.)

Was the resin getting to the end of its outlife? (Or had it maybe been re-lifed?)

How often was the plate debulked during layup (every ply or two is advisable for test pieces)?

How rough are the specimen edges (a polish with maybe 400 grit is usually a good thing)?

A couple of things which might give clues as to what's going on:

What is the nature of the failure surface (all within the resin or are there bare fibres)?

Are the failures true pure shear failures or is there some specimen bending going on?

-RP.

 

[SWComposites]

Other things to consider, all of which can affect ILSS:

- excessive resin bleed
- change in fiber sizing or surface treatment (don't count on the fiber supplier telling you about a change)
- change in resin ingredients or change in prepreging process parameters (temperatures, line speeds) (again don't count on the prepreg supplier telling you about a change)
-change in degree of impregnation (is the resin fully distributed thru the fiber tows)
- change in weavers or weaving process
- change from twisted to untwisted (or never twisted) tows, or vis versa
- prepreg out-time (time out of the freezer until cure)
- change in specimen thickness - you say the thickness is "usually" 8 plies. ILSS is a function of thickness (and a bunch of other things); it is NOT an invarient material property
- change in test setup (loading span, etc.)
- general random scatter: the test method is notoriously variable.

Steve