Problem with Aluminum expansion and dowel pins.

I have made some parts out of Aluminum for a customer. Pressed in 1/8" dowel pins, 3/8" deep.
Some came loose, found out later the part is a heater block. I used a .1245 reamer leaving .0005 for a press fit which seemed to press ok, no problems. Customer has no idea of the temperature at which the machine runs, it's just plugged in and goes. I am making replacement parts now and would like to avoid this problem in the future. Obviously 1/2 a thou is not enough, What size should I make the holes this time? The expansion rate of Aluminum is supposed to be .0000126 per degree, assuming 400 degrees means the part would change .005. Is that figured correctly? I am thinking of heading to his place with a Digital thermometer to get an idea of temperature. Any advice?

Can you use a threaded pin? Cut #12 screw threads, and leave small unthreaded area at the top and possibly a small pilot at the bottom to locate it?

You'll probably have to go to a threaded pin to keep it from loosening and that may not even work. Have an old rotary engine Suzuki bike that came with a finned aluminum exhaust manifold... real dumb idea. Rotary exhaust temps run even higher than reciprocating piston engine temps due to the lack of valves, head porting and other things that usually soak up a little heat. The pipe bolts work loose almost immediately. You keep re-tightening until they strip to the point you can't use them, then helicoil. When the helicoils come out, you have to have it welded up and re-tapped... then repeat the whole process. I'm sure that is operating at a far higher temp and much thinner sections that what you are dealing with, but just an example of how expansion in aluminum can bite you.

A potential problem is that an interference fit may cause plastic deformation of the aluminum at which point with sufficient heat, the fit will still get loose. Depending on the geometry of the part, the hole may get bigger, stay the same or get smaller with heat. If the part is like a donut and unconstrained, then the hole will get larger

Temperature and thermal expansion

Tom

Can you use a pull-dowel and use a small screw from the back side. It will have to be setup as a shouldered bore for the pin to "grab" something but you could probably come closer with a mechanical fix. Red loctite might last depending on the temps or it might run out like goo.

Could you use a roll pin or a groove pin?

"The expansion rate of Aluminum is supposed to be .0000126 per degree, assuming 400 degrees means the part would change .005. Is that figured correctly?"

Sorry, that's not correct. The coefficient of linear expansion is expressed as linear change per unit length per degree of temperature change. Here the units would be inch/(inch *F). A one inch piece of material would expand or contract its coefficient dimension if it underwent a one degree change in temperature.

I doubt if your part is pure aluminum, common aluminum engineering alloys coefficients of linear expansion vary from .0000128 to .0000132 inch/(inch *F). Lets say it's 6061 for example and use 0.000013. The expansion of your hole would be the change in temperature times the diameter of the hole times the coefficient of linear expansion.

The change in temperature from your shop temperature to 400 *F isn't 400 *F. Let's use 68 *F, a standard, for your shop. The change in temperature is 400 - 68 = 332.

The calculation of the change in diameter of the hole is simple: 332 X .125 X .000013 = .00054 inch

Don't forget the steel pin is also expanding so what matters is the differential expansion between the steel pin and aluminum hole.

Steel alloys vary also in coefficient of linear expansion, .0000065 is a good number for this first cut look at the problem. The differential expansion of the aluminum away from the steel pin is .000013 - .0000065 = .0000065 inch/(inch *F). Those numbers: 332 X .125 X .0000065 = .000027 inch.

Steel dowel pins are +.0002 over nominal: .1252, so if your reamed hole is truly .1245 there should be no loose pins at 400 *F with a .0007 press fit. How are you measuring the holes?

Reaming to tenths in aluminum can be treaturous.

I suspect part of the problem may be the hole was reamed a little oversize. Reamers aint perfect.even if new and you're careful.

A press fitted solid dowel in an aluminum part subject to thermal cyclin is a dubious design. A spring pin would be a better choice because of its built-in compliance. Use a generous entry champher on the movable part.

Aluminum Pin

For these circumstances I've used 6061-T6 pins that are Type III hard coated. Steel can be made to work with judicious tolerancing. Be sure to adequately countersink or counter bore the press fit hole or the material will bulge around the fitted pin and will not be flat around it possibly not allowing the mating part to clamp up properly.

Had something similar ages ago, cured the problem by reaming the dowel hole, then drilling & counterboring the dowel down it's axis for a tiny allen cap head tapped in to the ally.

Wasn't cheap but it cured the problem.

Rapid heating and cooling of the ali will cause the pin to loosen (steel dosnt change dimension as fast as ali)

Can you use a striper bolt with the head cut off? (would be able to save the part)

DaveE907 said:

"The expansion rate of Aluminum is supposed to be .0000126 per degree, assuming 400 degrees means the part would change .005. Is that figured correctly?"

Sorry, that's not correct. The coefficient of linear expansion is expressed as linear change per unit length per degree of temperature change. Here the units would be inch/(inch *F). A one inch piece of material would expand or contract its coefficient dimension if it underwent a one degree change in temperature.

Click to expand...

The dimensions of the thermal expansion coefficient are correctly stated as 1/delta t. The inch/inch furlong/furlong cancel out.


Anyway. Would aluminium pins have enough strength for the job?

It's just locating pins, 1/8" pins, 3/8" deep, .075 stickout. Had thought about threading. I used a new reamer held in R-8 collet, but have no way to accurately measure the hole, the press fit seemed fine. Will try a roll pin because I don't have to change the hole diameter, if that is not satisfactory then I'll drill and tap the hole. Thanks everyone for the help, Mark.

Just wondering what is so important or critical about locating this part. Does it really need the dowel pins?

Dowel pins locate a bracket that supposedly came loose and lost position.

"Rapid heating and cooling of the ali will cause the pin to loosen (steel dosnt change dimension as fast as ali)"

Yup, precisely. Used to take out seized up stainless plugs in aluminum pump housings by heating the plug red hot and packing the surrounding housing in ice. Again, an extreme example, but the result was that the plug expanded more than the aluminum and enlarged the hole in the aluminum. Aluminum also weakens at elevated temps (hot short), even before it melts. That's the reason the SR-71s had to be made out of titanium. Skin heating would anneal and weaken an aluminum structure at speed.

DaveE907 said:

Reaming to tenths in aluminum can be treaturous.

Click to expand...

But not nearly as treacherous as spelling.

~TW~

Thanks for the heads up, missed it after yet another 18 hour day.

Maybe you're looking at the wrong thing

I have been machining aluminum and pressing in pins for about 35 years, and I see a red herring in the last response about what the dowel pins do. If there is movement in the bracket, or too much load on it, it could be simply hammering the pins loose due to vibration. If the material allows, what about using larger pins, for starters? And make sure the bracket is securely fastened and not moving around. Perhaps, due to thermal expansion and contraction, over time, the screws or whatever fastens this bracket are becoming loose enough to allow the bracket to move. Perhaps the fasteners aren't sufficient for the pressure/load on the bracket. If the material of the bracket is sufficiently strong, the pressure/load would be applied to the weakest part, which, IMHO, is the pair of puny dowel pins. .075 of "stickout" is barely enough to actually engage the bracket in a meaningful way if there's the typical spherical radius or even a chamfer on the end of the dowel. If you really must stick with the 1/8" dowels, try making them longer, if it is possible. Another idea, use more dowels or more fasteners, or both.

Silver solder

mark costello said:

I have made some parts out of Aluminum for a customer. Pressed in 1/8" dowel pins, 3/8" deep.
Some came loose, found out later the part is a heater block. I used a .1245 reamer leaving .0005 for a press fit which seemed to press ok, no problems. Customer has no idea of the temperature at which the machine runs, it's just plugged in and goes. I am making replacement parts now and would like to avoid this problem in the future. Obviously 1/2 a thou is not enough, What size should I make the holes this time? The expansion rate of Aluminum is supposed to be .0000126 per degree, assuming 400 degrees means the part would change .005. Is that figured correctly? I am thinking of heading to his place with a Digital thermometer to get an idea of temperature. Any advice?

Click to expand...

Why not touch them with silver solder. Fast, easy and will not come out. There is also no need for a press fit.