Aluminium problems with heat transfer/bending

[andy21]

I have an aluminium (1050) `C` channel 100mm wide x 60mm high. It is formed from a flat sheet. The sides are at approx 30 degrees, so it is not a true `C` section but more of a trapezoid shape. Thickness is 5mm and length approx 2.2m. The graining runs along the length. It is also hard anodised. I need to transfer a generated heat through the 100mm face, up to 250 deg C. A heater mat is placed onto the inside face of the 100mm area, with self adhesive and is 80mm wide and finishes 50mm from either end. (a belt runs along the outer face perpendicular to the length - hence hard anodise - which in turn transfers the heat to the product running throught he machine). With the heater set at 200 deg C there is a bowing of the channel from the centre of approx 4-5mm along the length. The section is not fixed in any way that would restrict heat expansion and cause the bowing. This method has been used on my machines for years, but I need to run at 200-250 deg C now so the bowing is more noticeable than the lower temp used before.
Can anyone help with suggestions? I have thought about adding a plate to the `open` side of the heater mat to clamp it down and try to counteract the bending effect. But what material would this need to be? Stainless steel would be good for heat reflection but would the difference in heat expansion accentuate the bending? Also, if it was Aluminium would it need to be hard anodised too?
Are there and better grades of Aluminium that would resist the bending effects caused by the heating? An extrusion would be out of the question at the moment. I have looked into that and the quantities required would run into £1000s, not a good idea for an untested design.
Could it be that the heaters do not heat the ends (100mm shorter than the ALuminium length)?
Any ideas/suggestion would be greatly appreciated.

Regards,
Andy

 

[coughdrop]

Andy,
I work with Aluminium alloy castings rather than primary Al sheet etc., however the principles are the same. The 60mm sides are not heated, so the heated 100mm base is trying to expand more than the 60mm sides - hence the bowing. There is a Coefficient of Thermal Expansion figure somewhere, but it is specific to the Al content and varies with temp' bands. (for alloy LM0 99.5% Al = 0.0000285 per deg'C at 20-600deg'C) If your machine can tolerate it, why not utilise Al thermal coductivity better and INSULATE the inside of the 60mm sides? At least it won't incurr the costs of more heaters, or the energy consumption and it will probably reduce (not remove) the bowing to an acceptable level.
Hope this helps,
John.

 

[chicopee]

One explanation is that the bowing is due to uneven expansion of the aluminium. The side with the heat input is expanding more than the side with no heat. I would try to insulate the channel in order to maintain uniform temperature throughout the cross section and length. Another possiblity is from external loading on the channel suggesting that perhaps the channel needs to be beefed up.

 

[andy21]

Thanks guys.
There is no external loading at present (bench testing so far) so the bowing must be from the other theory.
I am now currently in the process of adding the heaters to all faces (they are not too expensive). I will also add a `lid` to the `C` channel to contain the heat as much as possible.

Regards,
Andy

 

[kenvlach]

This is intuitive rather than calculated, but might work:
When cool, bend a bow in the opposite direction. Then, when the channel heats up, it will straighten out.