[the4thseal]

Yes they would. Take it home. It needs a field trip.

[jsheerin]

As I'm sure you know, I guarantee you it will be a lower frequency than that if you haven't included the compliance of the linear bearings and ball nuts in your fea.

You could try an accelerometer if you really want to measure it. You could probably just run the input into a computer sound card and use a free real time FFT program like JDFT to look at the spectrum.

[danande]

Zach,
Nice pictures. Do you have any more videos of it in action with flood cooling?

Why would you want to stress relieve the steel?
Can the steel move after a certain amount of time, say a year, when it has been welded?

I've worked at a company that made supports and pipelines for offshore. We heat treated pipes and supports to remove any hardening, or brittleness that could have happened after welding, making the metal more ductile. Also to reduce potential hydrogen crackings, and stress corrosion cracking.

I know that you stress relieve welded steel if you're going to machine it afterwards, since the stresses locked in the steel can warp the metal when you remove metal.

PRE heat treatment on the other hand is a good idea.
Removing moisture and contaminants will reduce risk of hydrogen cracking and porosity in the weld. It will also reduce the stresses since the temperature difference is reduced.
If you get the temperature of the metal high enough before welding, combined with the extra heat from the welding, you will get a post heat stress reliving treatment with aircooling (to some degree atleast)..

Dan

[Zach_G]

I'd take a video of flood coolant except my mill has been crated up for the past couple months while I relocate for a new job in Houston.

Correct, in my application the purpose of post weld stress relieving is to prevent the metal from relaxing over time after it's been aligned. I've lucked out so far that mine hasen't moved in the past year, but if I were to do it over again I wouldn't take that chance.

The structural application of steel in a machine tool versus piping and supports is a bit different. In designing piping and supports it's important to max out the steel's yield strength (with a safety factor of course) so that the metal and cost needed to withstand the pressure and loads is minimized. Eliminating stress concentrators with proper welding processes prevents parts from fatiguing prematurely. In a machine tool the structure is designed to such a high rigidity spec that the stresses involved are practically negligible compared to the fatigue limit which is why I'm not too concerned with my crappy welds cracking and breaking =p

[danande]

I understand.
How would you preheat and stress relief the metal?
I've read that structural steel (s355) should be preheated to 200c, and stress relief at 600-650c. They say to heat the metal 1 hr per inch of steel.. So in my case ( 1/2'' wall thickness) its 30mins with 600c.. I don't think it's going to be easy to heat up the steel to 600c with a propane torch.. Preheating to 200c shouldn't be any problem.

Dan

[Zach_G]

Stress relieving most steels: heat to 1200F which lowers the yield strength to allow plastic deformation as a result of the residual stresses and hold at temperature for a couple hours. Longer allows more stresses to relax but there is a limit of diminishing returns. I misspoke earlier, you may air cool. Stress relieving is different from annealing which does require very slow cooling and also reduces the modulus of elasticity (rigidity) of the metal, not so good. Unless the metal is protected from atmosphere (vaccuum, inert gas, stainless foil wrap) you'll need to remove oxidation. I had my tubing stress relieved prior to welding by a company that does a lot of industrial weldments. They charge by the pound so was only like $80 for treatment and sand blasting.

[Zach_G]

Long time no post. Looks like most of my old posts/pictures were trashed in the update to the new forum, bummer. Hope the mods can recover them. Seems like theres a few other bugs from the migration.

Thought I should share a status update. Made some improvements. Most notably a half-decent spindle lock, works much better than a strap wrench. Basically a steel rod that engages a keyway in the mini mill spindle shaft with a captured spring that disengages it. Also, I got flood coolant now! Just a lil pond pump and some loc line. The drain is stuffed with stainless steel wool to keep out the majority of the chips. The sump is a 5 gallon bucket with a divider plate, any chips settle to the bottom as the coolant spills over through a brillo pad in there to strain out the floaters. Heres a video chewing on some annealed S7 tool steel. 4 flute carbide stub endmill. Toolpath is trochoidal to maintain an even chip load with 3/8" depth of cut, 28IPM, 2300 rpm and max 0.05" step over. Interestingly running coolant seemed to reduce the noise, but left a worse surface finish than dry running on steel. Definitely necessary to clear the chips and keep the spindle from overheating. Need to get a compressor and try an air blast.

[Zach_G]

My wee mini mill spindle has developed some rather excessive runout. About 0.0025" at both the end of the internal R8 taper and external dia. When new, it was only about 0.0005". At first I thought maybe the deep groove bearings were toast, it's only got about 60 hours of spindle time but I keep the preload pretty heavy and the bearings run hot. After installing new angular contacts there's no change in runout. Now I'm thinking maybe the spindle nose got bent (heck even the housing is pretty flimsy) by one of my many G-code follies as runout is only about 0.001" closer to the nose support bearing. A new spindle is on its way from little machine shop but I'm wondering what's the best method to detect a slight bend on a shaft with multiple differing diameters in close proximity?

I've also considered attempting an in-situ taper regrind something along the lines of Regrinding a mill spindle. Apparently that worked using a cheap HF pencil grinder... Since my mill head doesn't tilt, the grinder could be mounted to a spare linear bearing which is angled in a vice and indicated to match the r8 taper. Fun project but seems like a lot of work, hopefully the new spindle does the job.

[wizard]

My wee mini mill spindle has developed some rather excessive runout. About 0.0025" at both the end of the internal R8 taper and external dia. When new, it was only about 0.0005". At first I thought maybe the deep groove bearings were toast, it's only got about 60 hours of spindle time but I keep the preload pretty heavy and the bearings run hot. After installing new angular contacts there's no change in runout.

Well that sucks.

Is the spindle "soft"? It could be deformed from worn collets or tool holders. I didn't look back through the thread to see where the spindle came from.

Now I'm thinking maybe the spindle nose got bent (heck even the housing is pretty flimsy) by one of my many G-code follies as runout is only about 0.001" closer to the nose support bearing. A new spindle is on its way from little machine shop but I'm wondering what's the best method to detect a slight bend on a shaft with multiple differing diameters in close proximity?

Indicate the outside of the spindle as well as the inside. If the runout on the exterior matches the runout on the interior then you possibly have a bent spindle.

I've also considered attempting an in-situ taper regrind something along the lines of Regrinding a mill spindle. Apparently that worked using a cheap HF pencil grinder... Since my mill head doesn't tilt, the grinder could be mounted to a spare linear bearing which is angled in a vice and indicated to match the r8 taper. Fun project but seems like a lot of work, hopefully the new spindle does the job.

You will want a rigid set up for grinding the spindle. You also want the taper ground to a high degree of precision. It may be better to find a local grinding shop that can handle such work. I say that because I have no personal desire to stand on my head to get that grinder lined up right on a small mill.

[Zach_G]

It's an X2 "mini mill" spindle, supposedly hardened and there are signs of discoloration from heat treating on the non-ground surfaces. It has an alignment pin so collets always go in the spindle with the same orientation which shows up as 3 lines of wear where the majority of the pressure acts on the collet's "fingers". Runout is the same at both internal and external surfaces, and improves when measured closer to the support bearing so yeah kinda sounds like it got bent. I bet a grinding shop would charge far more than the $60 replacement from Little Machine Shop (Grizzly sells it for $32 but only as MT#3 taper). Hopefully with a new spindle the problem goes away but I do want to know the root cause to avoid repeating the same mistake. Could be fun trying to DIY a regrind, though that mentality is how I wound up building a CNC mill rather than an easier, cheaper conversion of an existing mill!

[clmenz]

Sent from my SAMSUNG-SGH-I317 using Tapatalk

[Zach_G]

Hello Zach_G,

I am in the process of rebuilding a mill and I came across your thread and your joint replication method. I am at a point where I am addressing the joining of the column to the base and I was wondering if you might share the materials you used for your filler. You mentioned epoxy/Iron Oxide in a 20/80 ratio by weight. Could you provide the name of the epoxy and iron oxide you used? How has this held up for you thus far? Thanks

Mike

Hey Mike, I used US Composites 635 Thin Epoxy with 3:1 medium hardener and mixed it with as much iron oxide as possible until it got too viscous to inject using a plunger. That was about 20/80 ratio by weight which yielded a mixture similar to honey. There's thinner epoxies out there which would allow a greater ratio of iron oxide but the US Composites is probably one of the best values. Just make sure to use an epoxy rather than other resins such as polyester since epoxies have the lowest shrinkage after hardening. The iron oxide powder is black Fe3O4 which is readily available on Ebay and often sold for making thermite. The joint has held up over the years with no discernible movement despite many crashes and some pretty rough abuse hammering on tool steel with indexable cutters. I highly recommend it!

[Azalin]

@Zach_G

Very nice build.

The Linear bearing blocks of X and Y are overlapping so how did you manage to screw them down?

Again, very nice project.

[Azalin]

umm, You've used long type blocks on the X?

[Zach_G]

An earlier version had the xy bearings bolted to the same saddle plate but as you noticed accessibility to the bolts was very restricted and difficult to work on. I modified the design so the xy bearing blocks bolt to separate plates that then bolt together along the exterior perimeter.

[shedbob]

Hallo Zach_G
Nice machine.Thanks for sharing your column epoxy recipe,I might use that in the future,nice to know it was successful over time.
Did the new spindle cure your problems? Did you ever determine the cause of the spindle problem?

[Zach_G]

Yep, new spindle fixed the runout problem, must've bent the old one while hammering on tool steel with a big indexable cutter. Guess my machine just doesn't have the rigidity for that kind of work. While I was in there also replaced the spindle deep groove ball bearings with some cheap angular contacts and wow what a difference! It can run 6000rpms now and the housing only gets warm to the touch rather than scalding hot, plus rigidity seems to have increased a bit also - a highly recommended upgrade for anyone running a mini mill spindle.