What is the best aluminum for heat sink?

formatting link

Reply to
Spehro Pefhany
Loading thread data ...

I guess if you can rotate a block on a flat surface by 180 degrees and it still matches a reference surface then it's square.

Reply to
Spehro Pefhany

I used to hang around an optics company that had a bunch of the slow wet grinding tables that made things flat. There is also an inherent (no external artifact) way to make perfect cubes, but I don't recall what it is. That shop was making the retroflector optics that one Apollo mission left on the moon, in a grungy shop behind the River Rondezvous bar at the foot of Carrolton Avenue, right near the levee, in New Orleans.

John

Reply to
John Larkin

Scraping has the advantage that it looks nice and retains lubricant on moving mating surfaces.

Reply to
Baron

Thanks, that solves one mystery. I'm still curious how the original right-angle standards are made.

I guess you could reflect a laser vertically off a pool of mercury until it returns perfectly back to the source, or maybe somehow rub three different cubes together in rotation while sliding them on a surface plate.

Best, James Arthur

Reply to
James Arthur

I'm thinking a plumb bob hanging down into a pool of water would make a pretty square reference angle, or maybe bounce a laser off a first-surface mirror so that it returns to the laser...

Hmmm.

Not that I need such precision. I've got the CNC bug[1], and I'm thinking of ways to make things nice. Ways to make nice ways, if you will.

[1] A CNC mill and a lathe, and the desire to make more of the same.

Cheers, James Arthur

Reply to
James Arthur

Good idea--optics people know all sorts of magic tricks like that.

Thanks.

James Arthur

Reply to
James Arthur

Very informative. Thank you.

Reply to
Bob Larter

That would've been a pig of a job. How long did it typically take?

Reply to
Bob Larter

I wonder if my uncle (recently retired old-school watchmaker) remembers. They had to produce two identical cubes such that any two surfaces would match in size/shape with detectable (by hand) stiction... using files and scrapers.

Clifford Heath.

Reply to
Clifford Heath

The lathe and mill are both Turing-complete machine tools, so that's no problem. :-)

Crossposting to RCM, which I recall had someone madly interested in something like "Godel's Theorem for Machinists"!

Tim

Reply to
Tim Williams

Lap all four faces independently of each other, then lap the block against one face, clean it, and mate it, and then attach the other face, being sure to accommodate any non-paralell circumstance the block may have incorporated into it. The second item mentioned would be the heat source.

Reply to
SuspendedInGaffa

Surface grinding for large mating faces that then get lapped together against a third, thinner lapping skate.

Ideally, surface grinding on a medium sized mating face, subsequently lapped.

Definitely should only need lapping on small mating situations. If the item is that far out of flat, something is wrong with the way it was made.

There is no way that "scraping" would yield true flat surfaces. His situation HAD to use huge amounts of gap filling mating paste of some kind. Scraping is NOT a precision machine operation. Anything bigger than a couple square inches should get surface ground. Anything below that size should be manufactured with a very flat face to begin with.

Lapping is certainly better for zero gap magnetic elements like transformer cores, etc.

Heat sink mating would benefit from lapping, but only if one ignores the labor cost. If you are doing a personal or proto project, sure, but if you want design for manufacture, you need to design it into your proto stages as well, and lapping is a VERY labor intensive, semi-skilled procedure.

So, ideally, in a manufacturing environment, reasonably integral, highly reliable matings are already being done every day without lapping.

If it is to be a non-permanent installation, which has service expectation or provisions to consider, the mating paste has to be one that does not form an adhesive bond.

If it is to be a permanent, one shot per part, per sink assembly, then the use of the Silver filled epoxy that the IC chip industry uses to bond chip tops to dies with is the right choice. It forms a strong, adhesive bond on the two faces, if they are clean, and has the best characteristics for permanent attachments which require good thermal transfer across that boundary.

Reply to
SuspendedInGaffa

I made a power supply (amplifier actually) for a Piezo stack that drove a tool head on a lathe at 20 - 20,000 Hz.

It could take a mounted round pin on the lathe, and cut it into a square pin with a pyramidal nose while the lathe spun at several hundred rpm.

It was used to cut contact lenses for thoroughbred horses. They were about one inch in diameter without the need for further optical polishing.

formatting link

Reply to
SuspendedInGaffa

The thing with a right angle defining device (a square) is that they self prove their own accuracy.

There is no way that a square that was off by even a TINY fraction of a degree could not be detected by any person with a good eye.

Reply to
SuspendedInGaffa

A block placed into the chuck of the lathe, can have a pretty damned flat face cut onto it, and a subsequent hand held lapping can be accurately performed on it. That makes one face true, and finely polished.

Then, it can be milled on the other face with a surface grinder or mill. A vertical mill would be nice as it would achieve a better surface quality than a horizontal mill would, and it would be kept more parallel to the bottom face. Back over to the lathe to mount up, and it should lap as true as the other face, which is a test of your upchucking capability. :-)

Now, you have two parallel faces, and if you mounted it into the lathe very carefully, the sides will be perpendicular to those two faces.

You could then continue upchucking different faces of the cube and facing them off with the lapping device you'll want to make for the tail stock of the lathe. :-)

I agree... upchucking is fun.

Reply to
SuspendedInGaffa

Very cool stuff.

Cheers, James Arthur

Reply to
James Arthur

pythagoras' theorem.

Reply to
Jasen Betts

You could always play with material addition rather than subtraction. There is an open-source rapid-protoype machine at

formatting link

Of course, you'd still have to buy some simple structural pieces and threaded rod (and raw plastic)... --Glenn Lyford

Reply to
Glenn Lyford

Scraping mostly went out of use around 50 years ago but is still ocassionally used, and can produce a flatness of .0001 inch over a surface spanning hundreds of inches, with a surface finish of about .0005 inch, well suited to holding oil. Skill required.

(Modern hardened way machines are ground, not scraped, and most machines are never reground after manufacture, they are just demoted to lower precision work as they wear out, or just scrapped and replaced with a new machine.)

Lapping has been the ultimate method for producing accurate surfaces for hundreds of years. I once read Isaac Newton's description of the process in an optical fabrication book, a section included to show how little the process has changed. (Newton did not invent lapping, but he was credited with being the first to write a description.)

Lapping can produce flatness of better than .000001 inch (one microinch) over many tens of inches, with RMS surface finish better than one microinch. I don't know what the largest optical flat ever produced was, but telescope mirrors have been produced by lapping with geometry and surface finish with microinch accuracies over hundreds of inches.

So lapping done by an expert is about two orders of magnitude more accurate than scraping done by an expert. Alas the process is little understood outside of the optical fabrication community, with people attempting to use totally inapropriate surfaces like a piece of glass for a lapping plate, without even making it flat before and during use!

(I have lapped parts up to 40 inch OD flat to better than 10 microinches with 1 microinch surface finish in a decades ago job as a lab tech; smaller parts flat to 1 microinch. Most of the work is in the shaping of the lapping plate, which needs a very slight crown to produce microinch flatness on the workpiece.)

Typically a surface can be economically milled or otherwise machined flat to about .001 per 10 inches, ground to .0001, and lapped to .00001, at high volumes in a production shop without undue effort. Tighter tolerances cost more.

BTW there were many variants of hand work exercises for producing accurate surfaces in the old days, mostly ending with WWII. One shop I know of required apprentices to saw off a piece of 1-1/4 bar with a hacksaw (sawing 3 sides) then file it to a 1 inch cube, with thickness and squareness +/- .001 inch, using a micrometer, a machinists square and feeler guages. If the foreman found any spot outside tolerance, usually a spot where the .001 feeler guage would slip under the square, then the apprentice got to start over again. No lapping or scraping permitted. A number of these hand work exercises as well as powered machine exercises were required for the promotion to Journeyman.

Reply to
Glen Walpert

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.