Butt splice or Overlap 3

[CSrac]

I have not found any imformation on the advantage of overlapping composite material when it is neccesary to splice. My company has a general spec that requires 1 inch overlap of woven graphite and unidirectional material. That is 1 inch overlap,3 inch stagger between plys. Why would butt splicing not be just as good if the splice was rotated 90 degrees between plys. Overlapping thick laminates leaves too much un predictable thickness build up opposite the tool side of the lay up. Anyone know of any written material on this sublect?

 

[BenGib]

If your dealing with relatively thick laminates you might want to consider a step or scarf joint, as far as I know its the best way at reducing stress concentrations and preventing delamination at run-out plies. I cant think of any reason at all why an overlapped joint would be advantageous with composite laminates. In any case check out Nyu's composite book, I seem to remember he touches on the subject.
Ben

 

[Compositepro]

I believe you are refering to splicing prepreg plies during lay-up and not splicing of the cured laminate. A butt splice will not carry any load accross the splice within the spliced ply. The load will transfer to adacent plies and across the splice. This is weaker than an overlap splice where the load transfers within the spliced ply through shear in the overlap.

Overlap splices are also much easier to make as there are very loose tolerances on the amount of overlap. Gaps are likely to occur when butt splices are attempted. In most applications it really doesn't make difference.

 

[RPstress]

Assuming this about ply splices, not joining whole laminates:

The overlaps are standard for strength reasons in fiber directions. However, for a UD (tape) material, there is no need for a ply overlap in the direction normal to the fibers. Typically, there is a limit on how far *apart* adjacent pieces of tape can be, to avoid gaps which resin would have difficulty filling (and maybe to keep fiber volume fraction from dropping). [NB: because UD material is usually supplied in great lengths on a roll, it is not normally necessary to have a break in continuous fibres, so an overlap splice is not usually necessary. It is woven material, with its limited width, which often needs a lot of splices.]

Note that butting plies, even in the fiber direction, has very little effect on laminate modulus. If it is stiffness designing the laminate, then the drop in strength may be acceptable.

For the splice size, roughly speaking, you have to be sure that the endload that is in the broken ply can get out of it in inter-laminar shear in the distance of the overlap (splice). Thus, the overlap is usually worked out in terms of a multiple of the ply thickness, with a chunk added on to allow for things shifting during cure.

For strong plies and weak matrix resin (e.g., UD carbon in wet epoxy at, say, 250 deg F), the overlap as a multiple of ply-thickness needs to be bigger than for weak plies and stronger resin (e.g., woven carbon in epoxy at RT). [It is the strength per unit width of the ply that determines the maximum possible load in it which would need to shear out. This is usually taken to be the ultimate allowable design stress times the thickness. The laminate's allowable inter-laminar shear strength should also be available as a design allowable.]

Similar arguments apply to the distance between ply drops, which usually affect UD material much more than splicing.

The stagger (I'm assuming this is the minimum permissible distance between splices) is more a function of common sense than it is justifiable numerately. If you put joints too close together, then the unbroken plies wind up with the little kinks in them (where they hump up and down over the splices) closer together; too close and you might get an unexpected problem and maybe poorer consolidation.

Where you cannot tolerate splices making a humpy bag-side (perhaps where a frame or other part must sit on a smooth surface), you need to either build up a sacrificial lump which can be machined back, or allow for the plies to be butted. The latter basically means making the laminate one ply thicker than it would otherwise need to be, so that the loss of the butted ply where it is broken is compensated for by the extra ply. This extra ply must of course be the same direction as the butted (broken) ply. For UD material that may mean one extra in each direction there's a butt (usually 0, 90 and +45 and -45). For woven material 0 and 90 are present in each ply, so you only need one extra in each butted 0/90 and +-45. If you have to line up the butts in more than one ply, you must add an extra ply for the "cut" fibers for each lined-up butt.

-RP.

 

[CSrac]

I was refering to the splicing of individual pre preg plys, particularly in structural parts with up to 15 plys. The problem I would like to improve is the bag size humps in the edgeband of parts that must mate and fair into the aircraft loft. Loft side is the tool side but ,but when installing parts such as wing to fuselage fairings, usually bag side of one is contacting tool side of the other.
Currently, when I use a butt splice on Woven prepreg, I am required to put a 2" uni cover across each splice, which defeats my purpose.
I like the idea of butt splicing, at least in the mating area of parts, then adding a full cover if necessary . It's more predidictable. Thanks for all of the input.

 

[Compositepro]

If you are only concerned about the edgeband area where parts must mate together you may want to consider a flexible caul sheet in these areas. It can be made from two or three plies of prepreg on your existing tooling using a rubber sheet to simulate your part thickness. This caul will keep your bag side surface flat in spite of overlaps. You will have a slightly higher fiber volume in the overlap as a result.