Measured Thermal Expansion of a 1.25" dia by 10" long Steel Bar in Response to Bare Hand Contact

[dgfoster]

This thread began as a response to questions/suggestions regarding my new right angle plate design. The concerns raised were based on sound theory but how that theory actually applied to the real world was a question I wanted to look at in a limited but geramine way.

Here are a couple pics of the plate milled.


I just recently milled my angle plate allowing me to set one up and answer the question raised as to how much casual hand contact with the strut would cause distortion of the plate. I wanted to be able to make a quantitative measurement of distortion rather than just blindly follow the warning to not touch the strut with a bare hand or, as others suggested, to make leather, plastic, or wood covers for the strut to insulate it.

Here are pics of the setup used.

You can see the base is a 12X18" granite surface plate. I 3-pointed the angle plate on the surface plate using small lightly lapped 2mm nuts. The setup is reasonably stable allowing grasping the strut without changing the reading by an interpolated readout a millionth. While grasping the strut the reading does change by a millionth to 1.5 millionths, but returns to zero on release. Lightly tapping on the base of the angle plate makes the needle jump but it does not remain deviated. Those observations indicated to me that the setup is robust enough to make reasonable measurements.

So, grasping the strut with an uninsulated hand does cause an easily seen and steady deflection of the face of the plate. It takes 25 to 30 seconds of steady grasp to deflect the face .0001" Upon release,the face slowly returns to its pre-contact location though that could take 20 minutes or so.

I'll be doing some more observation tomorrow though I repeated the above half a dozen times while doing some other work in the shop today.

For those doing super fussy work and wanting to avoid virtually all hand-contact thermal distortion, a simple garden glove or leather glove or equivalejnt will do the trick, at least on this casting. There was virtually no heat conduction with either a common leather work glove or a medium-weight rubber-palmed garden glove.

Because it takes so long to dissipate heat in the strut, a user should keep in mind that hand contacts are somewhat cumulative. In other words, holding onto the strut uninsulated for 10 second exposures repeated, say, five times over a few minutes would cause about 2 tenths deflection. Those same contacts spread over a couple hours would be of much less consequence.

So, I am not planning on covering up the strut or changing the design some other way. I will be casting another one (and a 36" camelback) in the morning if it does not rain.

Denis

I will be putting this measuring device, generously loaned to me by CarbideBob, to use more in the future looking at camelbacks (which I expect to be more sensitive) and other devices as my interest is piqued. It would be great to see other folks contribute some measurements of thermal distortion that they might make

 

[Bakafish]

Looks like a valid test setup to get the jist of this effect, and quite plausible results. That casting came out really nice, I'm impressed.

 

[dgfoster]

Looks like a valid test setup to get the jist of this effect, and quite plausible results. That casting came out really nice, I'm impressed.

Thank you. I am pretty happy with the casting. Since I cast the photgraphed casting, I have tweaked my core-making technique in a way that I expect to increase the crispness of the lettering. Casting is an interesting process that invites creative adjustments to potentially dozens of factors. I started out my first castings of this pattern using a simple match plate with loose pieces for the sprue and riser. I just finished permanently adding the riser, gating, and runners pieces as permanent parts of the plate as they have proven effective. I now am working on finding the optimal size for the core that will seal its print (making the border between the strut and the nameplate thinner and neater) without cracking the green sand and causing other problems.

Here is a pic of the riser pattern and the mortise I cut for it in the match plate.

And here is that riser and the sprue I use being printed on my Prusa filament printer.



Denis

 

[dgfoster]

Just to let folks look over my shoulder on a couple of observations made at 50 F, I made a pair of short vids. I think that makes the process seem more "real." The first shows a tenth deflection with a warm hand in a cold shop happening in about 7 seconds—pretty fast compared to a 65 or 70 deg shop.(My hand measures mid 80’s or higher) And then I made a vid for 40 seconds of the reequilibration rate (slow) after I removed my hand.

Hand contact at 50deg ambient after the device had sat at this temp over night:

I took my hand away and watched the needle crawl back toward zero.

Incidentally, just to test stability, I have left the checker on for the last 36 hours and noted less than 5 millionths (interpolated but likely very close) deviation from zero. There has been no heat turned on in the shop in that interval. I am surprised to see so little deviation.

In the next couple weeks I plan to set up a 36" camel back and see what it does at 70 F with hand contact. I will also probably load it with weight to confirm or deny prior deflection testing. The same is true for at least one or two of my prism designs.

Denis

 

[Gard]

Denis,
Excellent work, thanks for sharing.
I think little change as the shop cools down is reasonable A uniform lower temperature should not change the shape of a part.
It shows your electronic gauges are stable, sometimes they can drift a little with time or temperature.
David