O.K. Thanks Warren.  I guess I'll just have to try it out and if it isn't good enough then make a compensator.

Russell.

Hi All,

The problem with getting thermal coefficient data is often the manufacturers simply don't measure it.  It is not a commonly asked for bit of data with these kind of materials.

This is why generally you can only find generic figures for carbon composites.

This is even more the case with resin and raw fabrics as even if they could measure figures for each material, often they are meaningless when the material is only ever used in the form of a composite where the performance would be significantly different.

For our specific tubes, we have been asked about the thermal coefficients before and having checked with the manufacturer they don't have any test data.

Warren Penalver
Easy Composites / Carbon Mods - Technical Support Assistant

No, but it is usual to be able to quote a range of values as for all the other parameters given in the technical specification.  The most likely factor affecting the TCE would be the fibre to resin ratio and I would expect that to be closely controlled.

Regards
Russell.

I don't think you will get a reply. Do you expect the manufacturer of those tubes to test each batch?

I'm still hoping for a definitive answer from a staff member.  I guess that your tubes are bought in from an American company or a far eastern one as the published specs refer to ASTM standards rather than BS or EN so I understand if it takes a while to get the information.

Doing further research on the internet I have found specifications quoted from - 0.3 to +8 ppm/C in the direction of the fibres.  This strikes be as a very wide range.  I would like to get my completed pendulum length to within plus or minus about 0.2 with compensation if necessary.

TIA
Russell.

Thanks for your reply.  Yes, it is the longtitudinal expansion that interests me.  The period of a pendulum is proportional to the square root of it's length so I need to keep that as constant as possible.  The traditional material is Invar with the addition of a brass slug which lifts the weight to compensate for Invar's 1.3E-6 coefficient but I think carbon fibre composite could be better and cheaper.  (I've been quoted €267 per metre here in France for 6 mm diameter rod)

As I understand it the carbon fibres have a negative temperature coefficient and the epoxy resin has a positive temperature coefficient.  The coefficient of the composite will depend on various factors such as the fibre length and the fibre/epoxy ratio. Hence I am looking for the specification of Easycomposite's tubes.

Russell.

There are two different coefficients, one parallel to the fiber, one orthogonal (sp.?). The first is negative or pretty close to zero, the second is positive. The actual numbers are of course dependent on who you ask, every table gives other numbers... let's just say the first is -0.1*10E-6/Kelvin, the second is around 10*10E-6/Kelvin.

I am hoping to use carbon fibre composite tube for the pendulum of a precision regulator clock.  I believe that the coefficient of thermal expansion of pulltruded carbon fibre composite is somewhere between 0.1 and 0.3 ppm /°C but wonder if you can give a more exact figure  for your tubes so that I can design a compensator.

Thanks in advance.

Edit:  The coefficient is I believe negative.