Buckling of part of laminate. 1

[RPstress]

We recently had a failure of a test specimen which we couldn't explain, except for a mechanism like that sketched in the engineering.com picture. (Hopefully that works; it's listed as an 'attachment' in the preview.)

Has anyone got any opinions about the viability of the idea? A quick check on Euler properties doesn't rule it out.

(If crack is 25 mm long and four-point specimen is 2 mm thick total and 10 mm wide and QI HS carbon it might have Euler_stress = 4.π².50000.(1³.10/12) / 25² / (1.10) = 260 MPa average, maybe 500 MPa peak...)

However, intuitively it seems unlikely. Maybe if the curvature was not too great (flatter than in my sketch)? It looks like it needs a delamination to start with. Could this possibly happen with no delam?

The extreme plies under maximum compression seem to pop off in a failure mode similar to Euler instability, presumably modified by an elastic foundation effect due to being part of the laminate. This seems to happen to the extreme ply if it's at 0° and gives a buckle length of about 10 mm with no obvious fibre failure. Plies are about 0.25 mm thick.

 

[Compositepro]

The scenario you propose is very possible. A delam or simply a waviness in the fiber can initiate a buckle. The buckle will spread due to peel failure. A material with a low short beam shear strength will fail this way.

 

[RPstress]

Our material is a good Hexcel aerospace resin with IM carbon. Quite a good interlaminar shear as such things go, probably about 60 or 70 MPa for a multiangular UD or woven laminate. We can't see any obvious waviness and there shouldn't be any; these are quite carefully made test specimens.

How would this failure mechanism differ for a laminate all in compression, which doesn't usually appear to have noticeable buckles in layers? (Or not the type of large-scale Euler buckling I've condsidered as a possibility, anyway.)

 

[berkshire]

RP Stress,
From a practical point of view, this was the most common failure mode on fiberglass and carbon fiber sailplane tail booms. The peeling failure you are describing happened more often with unidirectional lay ups than with woven lay ups. This would always occur on the compression side of the tail. The cone would buckle and delaminate first, then tearing would start on the tension side.
B.E.

You are judged not by what you know, but by what you can do.

 

[blakmax]

Have you considered that the crack results in a region where there are effectively two unsymmetric laminates and the buckling is driven by coupling effects?

Regards

Blakmax

 

[RPstress]

blakmax:
The current situation is a bit complicated. We had a specimen in bending which has failed with a major delamination appearing about half way through. There is no known mechanism for this.

Some time ago there were three 4-point bend specimens with delams as per the sketch. Two of these failed in bending but a third failed with a delam extending as per the sketch. There was no known mechanism for this, hence it raised the possibility of some sort of instability, again as per the sketch. And if it could happen for that specimen maybe it could be a model for a failure mechanism which might conceivably occur for a laminate in bending without a precracked delam.

There is no known precrack in the test specimen that's failed, and although some such defect is possible we think it's unlikely. The current failed laminate seems to have failed in a somewhat asymmetric arrangement but not excessively so: typically about 5 mm (20 plies) in one 'half' and 6 mm (about 25 plies) in the other with repeated sublaminates of a non-symmetric quasi-isotropic layup which make up half-laminates which are pretty symmetric. The thinner half is on the compression side of the final delamaination. The precracked specimens from a while back that make the 'model' for such a failure (if it happened) were more symmetric either side of the precrack.

The current laminate is quite thick and shows some signs of compression failure of the surface plies in maximum compression, but the chief failure is a delamination as described.

I was curious as to if a failure involving instability of half a laminate is even possible even if a precrack exists; if so then I would extrapolate the query to the possibility of a laminate with no delam failing due to similar instability. That would raise the issue of compression failure involving instability of portions of the laminate, which is not a failure mode I've seen discussed anywhere.

Sorry to be unclear.

berkshire: could what you describe have been failure due to instability of part of the laminate?

 

[berkshire]

Rpstress,
This is most likely, these failures occur at well beyond the safe working load of the laminate. The scenario is that an aircraft made from FRP or CFRP will have an incident on landing, or sometimes during the takeoff roll, causing it to depart from a straight line. The resulting ground loop puts a side strain on the fuselage cone. The side of the fuselage cone that is in tension does quite well, however the compression side will buckle. If the load is removed before failure, the structure will pop back to its original state, with no apparent damage. if the structure is loaded to failure, then cracking of the outer layers and delamination between the layers is evident. Sometimes during the repair, delamination and white marking, where the fibers have de bonded from the resin will be evident up to 6" either side of the break. This delamination is much harder to see on CFRP and dye techniques have to be used to find it.
I realize I am rambling here, but its hard to give a yes or no answer.
B.E.

You are judged not by what you know, but by what you can do.