h a l f b a k e r y
Number one on the no-fly list

meta:

account: browse anonymously, or get an account and write.

 user: pass:
register,

Temperature Compensating Micrometer

Use bimetallic construction to offset thermal expansion
 (+8) [vote for, against]

The micrometer <link> is a device used to measure the outside dimensions of objects, at best they can be accurate to 1 micrometer. This is pretty impressive. They're available in a wide range of sizes and usually made of high-quality steel in a C-shape. Because they're compact & portable they're used in a wide range of environments. This is a problem, because micrometers are made of stuff, and stuff expands & contracts with heating and cooling. That means that the steel C-shaped backbone of the micrometer will expand, by roughly 10 micrometres per metre per degree C. Or 1-5 um/C for a typical instrument. Worryingly, that error is right in the range of what you're trying to measure. That being micrometres.

What you're supposed to do about that, is perform your measurements in a consistent temperature range. That's not always possible. Maybe you want to measure a component in a hot/cold environment to see if it's out of spec. If you take a cold micrometer into a hot environment and make 5 measurements of the same object, the readings will gradually decrease as the instrument warms up. This is not ideal. What can we do about this?

So, we change the shape of the tool from a C to a steel H. The measurements are between the lower two vertical parts. The cross piece of the H will be steel and subject to standard expansion and contraction. Between the top two vertical sections will be a solid cross piece of another metal. A good example is brass, which expands at ~ 2x the rate of the steel. That means that as the middle steel cross piece expands 1um, the top cross piece expands 2um and, with the central piece acting as a fulcrum, pushes the measuring surfaces between the two lower vertical sections together by 2um. Thus, offsetting the 1um expansion in the fulcrum. There may need to be some experimentation with the two metals, lengths of each section and mitigation of bending etc, but the principle is sound.

Now, you have a relatively temperature insensitive measurement tool that should be more precise in the field, even if that field is Antarctica, or space.

 — bs0u0155, Mar 04 2021

Micrometer https://en.wikipedia.org/wiki/Micrometer
[bs0u0155, Mar 04 2021]

Negative thermal expansion https://en.wikipedi...ansion#Applications
Maybe you could use a negative thermal expansion material [xaviergisz, Mar 04 2021]

 Problem is, now you have to also ensure that every component you are measuring is also made of the same bimetallic compound otherwise your measurements are going to be fluctuating in the opposite direction.

 Perhaps the solution is to have a whole set of micrometers, each with a different bimetallic layering to give a different co-efficient of expansion. Then all you have to do is to choose the appropriate instrument to match the material of the object you are measuring.

Or a micrometer with a built in blow-torch to heat up the sample, that could work too.
 — pocmloc, Mar 04 2021

 //your measurements are going to be fluctuating in the opposite direction//

But there's nothing wrong with *that* fluctuation. *That* fluctuation implies that the things measured are getting larger or smaller with temperature. And they are. Unlike the fluctuation which this idea seeks to correct, it's not an artefact, it's just a facty fact.
 — pertinax, Mar 04 2021

 Ok let's try another line of criticism

Your geometry is only going to give correct results at the maximum width of the opening. Because the screw also expands, this micrometer will be out by 1um when measuring very thin things, while a standard one will be correct (because the screw fully extended across the gap will expand just as much as the C)
 — pocmloc, Mar 04 2021

Ooh clever.
 — RayfordSteele, Mar 04 2021

 //Or a micrometer with a built in blow-torch to heat up the sample,//

 I'm specifically not correcting the sample. It's the sample in it's actual environment I want to measure. There's no use disassembling a machine and measuring a specific component when it's not under any load, and it's in perfect lab conditions. The information we want is: "Shaft c was subject to excessive wear because, due to high operating temperatures and a compressive axial load it's diameter exceeded spec."

 You could put a small restive heater in the micrometer and use that to compensate. With heat you can only compensate up, so the instrument should always be at something like a toasty 100C, which should discourage theft and rust.

 //Because the screw also expands,//

It does. I wonder if they take this into account, and maybe calibrate the things at half max opening to build in a fudge factor? I don't know, but I strongly suspect that a lot of knowledge about how this sort of thing works was gained in the early days and now companies that make them, just, keep making them the same way without knowing why.
 — bs0u0155, Mar 04 2021

 OK well that makes sense now that you have specified more precisely what you are trying to measure.

 Now I like the idea of maintaining the micrometer at a certain temperature. Make it 1000°C and you have just invented the all-in-one micrometer toaster. You can now enjoy freshly toasted toast which you know the thickness of to the nearest 1um. You could even measure the thermal expansion coefficient of toast, as it cools.

//fudge factor// Now that's the kind of thinking that got us into the pickle that you are valiantly trying to solve.
 — pocmloc, Mar 04 2021

 //Digital micrometer + built-in electronic thermometer//

The way to do this is probably through electronics, but that's way less fun.
 — bs0u0155, Mar 04 2021

[bs0u0155]; that ^ I like. Even Lego Technic is going the way of "just put motors everywhere and control it all with software" instead of cool mechanisms and gearboxes and stuff. Mechanical solutions are just BETTER (& easier to understand & fix...). Although I might be biased...

I assume you're planning to design the brass and steel parts so they change temperature at the same rate? Thermal conductivity of brass is about twice that of steel (or up to 10x for stainless steel depending on the alloys used). Maybe by carefully adjusting the thickness and adding fins to the steel parts, you can get the metals to change size simultaneously. Be sure to wear insulating gloves to avoid heating part of the device with your hand.
 — scad mientist, Mar 04 2021

 // Even Lego Technic is going the way of "just put motors everywhere and control it all with software"//

 Ugh... I know. It all started going wrong when they started with specific panels to increase realism... Great, My Lego Porsche does look a bit more like a Porsche, but 50% of the kit is useless for anything else. If I wanted something that looked exactly like a Porsche, I'd have bought a Tamiya kit or something.

//brass and steel parts so they change temperature at the same rate?// Thermal conductivity isn't a big factor here I shouldn't think. They're linked together and the main limiting factor will be the air-metal interface which will be similar. It's possible some compensation might be needed, but a thin coat of paint will probably be all that's needed.
 — bs0u0155, Mar 05 2021

Thinking about getting a 3d printer so I can print me off some custom lego pieces.
 — RayfordSteele, Mar 05 2021

If you're going to be 3D printing, you can make any shape, why bother with the intermediate Lego step? Besides, there's a lot more fiddling about with 3D printing than I imagined. A resin printer could get you Lego blocks that fit right more or less every time, but there's clean up involved. With FDM, you're likely going to have to mess around with and adjust the final part to get it to fit into other Lego blocks right.
 — bs0u0155, Mar 05 2021

When I was a child I had my grandfather's Meccano set (the earliest parts of it were his older brother's from before WW1 I think), and I always had this idea that there were Technical Lego people, and Meccano people. And of course I considered Technical Lego to be a kind of strange inferior plastic Meccano-substitute.
 — pocmloc, Mar 05 2021

 //kind of strange inferior plastic Meccano-substitute.//

I'm ok with both, although a look at the Meccano website reveals they've gone the same way as Lego Technic. A big difference for me was that the things you made with Lego could be played with for a reasonable amount of time, with Meccano, the nuts and bolts seemed to rattle loose all the time, the key was to add two nuts and back the first into the second. but that made all the joints clumsy.
 — bs0u0155, Mar 05 2021

 // all started going wrong when they started with specific panels// Agreed, the barrel on the technic truck 42112 is just lazy or penny pinching. Or, on second thoughts, a possible puzzle challenge.

No lock nuts in Meccano? tiny bit of plumbers tape?
 — wjt, Mar 05 2021

 //tiny bit of plumbers tape?//

Something, something DET CORD!
 — AusCan531, Mar 06 2021

After a half century or so of poor storage in an old cigar box, all the surfaces get slightly corroded and then everything grips very nicely.
 — pocmloc, Mar 06 2021

 I like it, but (without first reading any of the other comments or actually going to check) I'd be really surprised if they didn't already exist (or hadn't already been tried & found wanting).

 Used to do quality control work for a small engineering firm turning out car parts & the like on lathes in my teens, ashamed to say we didn't use to worry about this issue overly much.

 I imagine for important work the simplest & most commonly applied method is just to use a temperature controlled room.

& for anything really important a laser measure used at both extremes of the parts expected operating temperatures.
 — Skewed, Mar 06 2021

 //commonly applied method is just to use a temperature controlled room.//

This is essentially the definition of a lab. But You can't always take a component into a lab. My grandfather used to make & refit parts for ships for Cammell-Laird back in WW2. There are situations where you might want to measure something like a propeller shaft. This is a component that is subject to temperature changes, heating due to a bad bearing in one spot while the far end is in winter sea water. All the while, you need to make measurements in an engine room which can be extremely hot, or cold. It would be ideal if we can measure something like that prop shaft without disassembling the ship, AND if the measurement tool weren't thrown off by how hot the room was at the time.
 — bs0u0155, Mar 06 2021

Have a suitcase of micrometers calibrated for different temperatures in 1° increments
 — pocmloc, Mar 06 2021

Oops sorry, deleted that comment, it wasn't right, really doesn't matter that different materials being measured expand at different rates, that would be what you want to know after all, what the parts maximum & minimum width is at different temperature extremes, so you only need to correct for the expansion & contraction of the micrometres materials.
 — Skewed, Mar 06 2021

 Wouldn't the lab be smart enough to have one end in the cold and one end at engine room normal for baseline measurements? Normal operating conditions not just perfect fit together measurements.

So these measurements are for prediction of wear, catching wear? The relative data set should still show change.
 — wjt, Mar 06 2021

All that aside I'm not entirely sure an ordinary room with a working thermostat really qualifies as a lab ;)
 — Skewed, Mar 06 2021

There's a lot of DIY garage/bedroom sound labs now with vibration affected thermostats.
 — wjt, Mar 07 2021

 // it's not an artefact, it's just a facty fact//

[marked-for-tagline]
 — 2 fries shy of a happy meal, Mar 07 2021

 //I'm not entirely sure an ordinary room with a working thermostat really qualifies as a lab ;)//

It's a start. You need a few other things but most are specific to the experiment you're running. The idea is a controlled environment, a thermostat is pretty snazzy, historically speaking.
 — bs0u0155, Mar 08 2021

 I once changed the water in my brother's aquarium. I took the fish and put them in a big glass cup, and then decided that the water might be too cold for them so I put the "thermostat" in the cup and plugged it in. The thing was that the heating element was in the cup, but the metal thermostat arm was hanging in the air not testing the heat.

I was able to get the aquarium real clean ridding it from all that green algae (or however its spelled) when I looked up at his beautiful fish which where all floating at the top of the bubbling and boiling cup.
 — pashute, Mar 08 2021

 //I once changed the water in my brother's aquarium. I took the fish and put them in a big glass cup//

 Oh god, I could feel where this was going immediately...

You change SOME of the water, the fish can stay. If you really need to move the fish, just use one of you're spare tanks that you have already set up as a quarantine/hospital tank... we all have those don't we class? If we need cleaning products we use peroxide/bleach/vinegar as these can all be neutralized in some way. Soaps/detergents shouldn't be allowed in the same area code.
 — bs0u0155, Mar 08 2021

Maybe what is needed is a thermal response that slightly twists and untwists the threaded rod to shift the pitch of the threads with temperature. I imagine that a wire spring wrapped around a core might be able to act as a threaded rod that changed length but did not change pitch. The read could not be via the standard collett but by a worm gear possibly?
 — WcW, Mar 08 2021

Right, this is a better expressed version of what I was trying to say. So it is the screw which really needs to be compensated. And [bs] agreed and suggested that "fudge factor" be employed instead.
 — pocmloc, Mar 08 2021

 //The actual problem is with the ruler itself//

 I largely agree with your (& poc's point), but I thought up the mechanism first and because I liked it, started to think about an application that it could be shoe-horned into. This is a half-baked idea after all.

 Also, the problem IS with the ruler, and everything else. The ruler is subject to thermal expansion, but, the C/G shaped support structure is much longer and subject to some level of relatively increased temperature-induced distortion.

I suppose what you could use this mechanism for is to create temperature-independent length standards, but that's not as fun.
 — bs0u0155, Mar 08 2021

 //a wire spring wrapped around a core might be able to act as a threaded rod//

I love that idea, [WcW], but wouldn't you have a problem with the compressibility of the spring?
 — pertinax, Mar 08 2021

 You would have a bimetallic screw, a copper screw and a steel screw, perhaps concentric and running in seperate threads, and then at their ends they are attached by a lever and cam assembly (a bit like the initial idea frame), disposed such that the length from thread point to anvil point is maintained constant through a range of temperatures.

In other words, miniaturise this idea and fit it inside the screw
 — pocmloc, Mar 09 2021

 [annotate]

back: main index