Shrinkage of GRP Mouldings

[stuff]

Peeps!

My Design:
I am designing a structural GRP moulding which needs to have connection interaces with extremely high accuracy. The part is 3m long, 1.5m wide and 1m deep (10' x 5' x 3.5') and will be built by separately moulding key features and linking them with a set of laminated PU webs.

My Problem:
I understand shrinkage will be my biggest hurdle in trying to maintain accuracy. The most shrinkage I think I can accommodate is about 1mm on the length. I am considering having steel spars moulded in to resist shrinkage.

My Questions:
1. Does anyone know any good fabrication techniques for minimising shrinkage?
2. Would anyone like to hazard a guess at the likely shrinkage?
3. Does shrinkage worsen over the longer-term?
4. Does anyone know of some good web or book references that I should see?

Thanks in advance.
Hedge

 

What is the process and material you propose to use?

 

[GregLocock]

I think 1 mm in 3000 is a very high target for shrinkage. Our solar car halves had errors of about 2 mm.

The GRP structures will warp over time unless you cure them properly. We used epoxy and had to bake it at 80 degrees C to cure it. Exposure to Asutralian sunlight (full sun at 50 degrres) caused significant warpage in the flat panels after 5 days or so. The skin temeprature probably exceeded 60 degrees.

Try and talk to boat builders and other custom GRP fabricators. They do understand this stuff, they don't write books about it!

By PU do you mean polyurethane? It will be very difficult to bond that to a GRP structure, unless you use a PU resin for the GRP, which I've never heard of, but it should be possible.

Cheers

Greg Locock

 

[stuff]

Thanks for the responses.

The part is fabricated from a glass-fibre csm and polyester resin. I thought it was a polyurethane foam but I could be wrong. The mould is a combination of GRP and fibreboard.

Our vendor is pretty good at counteracting shrinkage but can only promise a build tolerance of 3mm

Thanks
Hedge

 

[GregLocock]

Our problems were temperature related - if you are very clever you can use Kevlar's NEGATIVE expansion with temperature to control your structure. We didn't bother, we moved the neutral axis of the skins of our parts towards the fragile layer of cells, with a reasonable degree of success.

If you can't get the dimensional accuracy you need from a single moulding then can you either post-pierce the critical dimensions, or build it from two parts and jig them together with a slip-joint?



Cheers

Greg Locock

 

[WKTaylor]

This topic may be better discussed in the COMPOSITES forum...

forum327