PDA

View Full Version : High speed machining with a high speed spindle



Evan
01-29-2008, 02:45 PM
I have tried several ideas to make a high speed spindle (HSS) for my CNC mill but was not satisfied with the performance until this third design. The issue is that I want it to do actual machining, not just engraving. There are significant benefits to having high rpms available when machining aluminum and there is virtually no practical limit to how fast it can be machined.

What I settled on is using an R8 to MT2 adapter that I bought for my mill but have never used, until now. I machined it to accept a pair of spindle bearings of good quality. I used an angular contact electric motor grade Timken bearing for the lower bearing and a deep groove SKF Explorer series for the top. I chose these particular bearings because I had them on hand but they are well suited to the application.

The following set of images illustrates the high speed spindle and how it is mounted to the machine. The base plate of the HSS is steel with 41 degree dovetails on the top and bottom edge. These dovetails engage four 1/4" flathead countersunk cap screws by wedging under the heads similar to a regular dovetail. Because of the geometry only one screw need be loosened to install and remove the HSS. It can be installed or removed in under 60 seconds with perfect alignment. Initial alignment is performed by shimming the mount plate to the housing.

http://www3.telus.net/metalshopborealis/pics4/banshee1.jpg

The spindle is designed to operate from the original drive belt with absolutely no change in adjustments. Because it is a serpentine belt it is easy to pry off the main spindle pulley and put on the HSS pulley. This only takes a few seconds to perform, in either direction.

The orientation of the drive belt can be seen in both circumstances in the next set of images. When I designed the head assembly I kept in mind the addition of a high speed spindle in the future. The idler pulleys are preexisting as they are used to control belt slap during operation. Since that adjustment is not critical they are adjusted to provide the correct tension for the HSS and will then also do the intended job for the main spindle without change.

The drive pulley on the HSS is sized to provide a little better than a 1 to 4 step up from the main drive motor. Since it is capable of turning at least 7000 rpm that gives a theoretical top speed of 28,000 rpms. That is too much for the bearings to stand on a continuing basis so I have throttled back the main drive to 6000 rpm. That results in a HSS max rpm of around 22,000 in practice and I am not running it over 20,000 for now. The bearings are rated for a minimum of 25-26000 rpm depending on grade of oil and cooling.

To verify the actual rpm accurately I used a homebrew photo tachometer. By pointing an infrared photo transistor at the spindle and placing 3 vdc across the collector/emitter it will switch on/off as the spindle rotates, once per revolution. This is done by shining an incandescent lamp high in infrared on the spindle and covering one half of the spindle with black electrical tape. The output of the photo transistor is connected to a good quality frequency counter. It can directly count pulses per second and so will give a reading in revolutions per second. Multiply by 60 and you have rpm accurate to +-60 counts per minute, plenty close enough for my purposes.

As seen in the below images the counter is reading 438 rps. That corresponds to 26,280 rpm, a bit too much for the bearings so I adjusted the maximum speed pot in the controller to keep it within a safe range.

http://www3.telus.net/metalshopborealis/pics4/banshee2w.jpg

I put the machine through it's paces the last couple of days and am happy with the performance, as I said. I did some experimenting with drilling holes in aluminum as I have some applications in mind. I want to make some special composite materials based on an aluminum honeycomb type of core material.

I set up a sample for drilling and tuned the machine for maximum performance. It is able to drill 1/8" holes in 1/4" 6061 t-6 aluminum plate with a cycle time of 2.9 seconds and a spindle speed of 18,000 rpm. It can maintain this indefinitely which results in a rate of over 1200 holes per hour. This is sufficiently quick to make some of my ideas practical. I want to make composites of aluminum honeycomb covered in either plastic laminates or thin aluminum skins. The first real test I ran was a small piece that turned out ok except I was slightly short on the depth. That gave me the idea of not drilling all the way through which then only requires a skin to be applied to one side.

I ran another test, this time a larger scale test with an array of 512 holes. It came out very well with only a few discernible mistakes that are attributable to the Cad/Cam draftsman. I shall have a talk with him about quality control and lessons learned in our next production meeting.

The test piece under the drill:

http://www3.telus.net/metalshopborealis/pics4/banshee4.jpg

The final result: 58% of the material has been removed with a resulting average density of only 1.08. The piece is still subjectively extremely strong even without any laminations applied to the surfaces. The piece is very light weight feeling and weighs only about 42 percent of what it did. This will be useful in the future for building very lightweight apparatus that must be ported about.

The final piece:


http://www3.telus.net/metalshopborealis/pics4/banshee3.jpg

I have made a video of the machine in operation with the high speed spindle drilling and milling the piece. I have a new server co-located with the primary provider so it should provide for faster downloads, I hope. If you download either of the videos I would appreciate knowing what rate was achieved.

Full hi res version, 15 megabytes, running time 3:22

http://www3.telus.net/metalshopborealis/banshee_hi.wmv

Lo res 320x240 version, 7.6MB, same length

http://www3.telus.net/metalshopborealis/banshee.wmv

A bit later I will put up a version for other operating systems.


Full hi res version, 7.2 megabytes, running time 3:22, MPEG4 format

http://www3.telus.net/metalshopborealis/banshee.mp4

Note: If you have Media Player Classic this will play on Windows machines.

Greg Parent
01-29-2008, 03:37 PM
I downloaded the full hi res version, 15 megabytes, running time 3:22

http://www3.telus.net/metalshopborealis/banshee_hi.wmv


and got a 546 Kb/sec rate.

Nice work.

ubooze
01-29-2008, 03:47 PM
I d/l the Hi res version at about 20 kb/sec, but I am using my school's computer lab.

Very cool stuff, but what about that other spindle when your doing some more dimensional pieces rather then flat stock?

ptjw7uk
01-29-2008, 03:49 PM
Nice one Evan.
Whats the main motor power output!
Will the belt stand all that bending for long!!
Peter

Evan
01-29-2008, 05:01 PM
You have a good point Peter and I will make a roller to go over the bearing today.

It's a 1.5 hp motor so with the up gearing it is basically running about 1/3-1/2 hp to the spindle


but what about that other spindle when your doing some more dimensional pieces rather then flat stock?

Nothing is perfect.

2ManyHobbies
01-29-2008, 05:42 PM
Heh, you know you might be in a hurry when your bearings require active cooling. Any faster and we'll all be asking about the gas turbine spindle you are running!

As a side note, you've got me wondering about a device that could lay down foil and spot-weld a honeycomb one row of cells at a time. Well, maybe a double row of cells back to back, but that'd be later. :rolleyes:

BobWarfield
01-29-2008, 06:02 PM
Nice work Evan!

Do you plan to rig up a manifold to delivered compressed air to clear chips?

Best,

BW

PS What feed rate on the milling ops? Any special end mills?

S_J_H
01-29-2008, 06:48 PM
Evan, that is a very impressive setup you built!
Is the spindle housing a bolt together deal or machined from a solid hunk?
What are you using for bearing preload, a threaded adjuster?

Steve

Evan
01-29-2008, 07:56 PM
The housing is all bolt together pieces of 1/2" and 1/4" plate. The preload adjuster is a bit different. It has an ordinary bronze nut on the top end but I only had so much length and especially thickness in the R8 adapter to work with. There wasn't enough meat to slot the top end of the spindle for a key. I slotted the pulley and made a close fitting threaded key instead. The thread on the spindle is a special since I was trying to preserve as much metal as possible. I threaded it 20 tpi. To key the pulley I made a close fitting threaded key that fits the slot in the pulley and engages the threads in the spindle. The threads go all the way down to the top of the bearing. This provides an interesting way to adjust and lock the preload. With the bronze nut off the key is put in place on the threads and the pulley slips on over it. It can be turned freely until the key, which is following the threads, bottoms out on the top of the bearing.

The key is made about .050" shorter than the slot in the pulley, equal to one thread pitch. Once it touches the top of the bearing the obvious bearing lash is taken out and the position of the key noted. The pulley is removed and the key is moved up one thread pitch. The pulley is replaced and the nut installed. Then, by adjusting the position of the key with the pulley and the position of the nut the backlash is taken out. As this is done the nut locks itself against the threaded key while also pressing the pulley against the top bearing, setting the preload. The pulley can't revolve on the shaft since the key is locked in place.

It works just fine and takes much less time to do that than to read about it.


Air to clear chips is likely. I already use it by hand. It would also help a lot to keep the spindle cool.


Feed rate on the milling was 500mm per minute I think. Or maybe it was 1000? I was using the override so I don't recall exactly. Those particular end mills were made in China. They are quite good. Another batch I have would do well to mill machinable wax. I tend to buy the better ones but don't want to break too many of them testing this setup. I have already broken several finding out what I can't get away with.

rotate
01-29-2008, 08:22 PM
Did you use any lubrication? In parts of the video it looked wet.

The feedrate when you're cutting the perimeter was pretty impressive!

Evan
01-29-2008, 08:56 PM
I used some ethanol for cooling mainly. I was being semi conservative on the feed rate as it was programmed to do three passes and I could have jacked it up. I didn't want to mess up the piece. I have cut the full thickness with an an American made Niagra end mill at the same feed rate but broken a Chinese tool trying to do the same. Most of the Chinese tooling lacks toughness.

Uncle O
01-29-2008, 09:34 PM
537kb/sec.....

Rookie machinist
01-29-2008, 09:41 PM
Thats some nice work evan. When you use the high speed spindle how are you controlling vibration? I was just thinking at thoose high rpms even the smallest unbalance in the tool or the collet would vibrate like crazy.

pntrbl
01-29-2008, 09:53 PM
You're boggling my mind again ........ :D

SP

tony ennis
01-29-2008, 09:57 PM
You're a machine, Evan.

210kbps

Your Old Dog
01-29-2008, 10:08 PM
quick! Somebody Stop Him.




:D

Evan
01-29-2008, 10:22 PM
The spindle was turned between centers to ensure absolute concentricity of the bearings. I used very sharp C1 carbide tooling to cut the hardened metal. I just finished it and my wife brings me home a Valenite ceramic insert. Gee, thanks honey. :D There is a very small imbalance that becomes apparent over about 20,000 rpm. I'm going to play around with a collar on the outside of the spindle containing several set screws that can be wound in and out to alter the balance slightly.

andy_b
01-30-2008, 12:41 AM
I have made a video of the machine in operation with the high speed spindle drilling and milling the piece. I have a new server co-located with the primary provider so it should provide for faster downloads, I hope. If you download either of the videos I would appreciate knowing what rate was achieved.

Full hi res version, 15 megabytes, running time 3:22

http://www3.telus.net/metalshopborealis/banshee_hi.wmv

Lo res 320x240 version, 7.6MB, same length

http://www3.telus.net/metalshopborealis/banshee.wmv

A bit later I will put up a version for other operating systems.


Full hi res version, 7.2 megabytes, running time 3:22, MPEG4 format

http://www3.telus.net/metalshopborealis/banshee.mp4 (http://www3.telus.net/metalshopborealis/banshee_hi.wmv)

Note: If you have Media Player Classic this will play on Windows machines.

i maxed out around 290kB/s. oh, the third link (the MPEG4 one) is the same as the first one (the wmv one).


very cool!!!!

andy b.

Evan
01-30-2008, 03:44 AM
oh, the third link (the MPEG4 one) is the same as the first one (the wmv one).

Yes, it is. Yet Microsoft, in their infinite wisdom, doesn't support it because the codec is NIH, Not Invented Here. It will play on Media Player if you have the correct codec from a third party. It might play on the latest MP11 but I don't have that.

HTRN
01-30-2008, 05:17 AM
It does indeed play on Media player 11.


HTRN

andy_b
01-30-2008, 01:23 PM
Yes, it is. Yet Microsoft, in their infinite wisdom, doesn't support it because the codec is NIH, Not Invented Here. It will play on Media Player if you have the correct codec from a third party. It might play on the latest MP11 but I don't have that.

you may have misunderstood. the file that the third link goes to is the exact same file that the first link goes to. both are banshee_hi.wmv. i didn't try just changing the file extension, but i'm sure that would get me the mpeg4 file.

i was just mentioning it. the mill is still totally cool!

andy b.

Evan
01-30-2008, 04:57 PM
Thanks for clearing that up. The #^%&* advanced mode on the post editor here hides that from you when you copy and paste from another previous URL as I did. Even if I edit the link to point to someplace else and it all looks good in the editor it doesn't change the actual target and doesn't let you know in any way. I forgot about that detail. It's fixed now.

macona
01-30-2008, 05:55 PM
You might be able to take the spindle down to a shop that rebuilds turbos and get them to put it on their dynamic balancer.

Also it could be resonance in the belt and not an imbalance as so much.

macona
01-30-2008, 05:59 PM
Looking at the video it seems your Z axis acceleration times are rather slow. Will it miss steps at higher accel?

Evan
01-30-2008, 08:32 PM
The Z axis is geared 5 to 1 so under normal operating conditions it's going as fast as it can reliably. It's already accelerating at 45,000 hz per second. Top speed is 35 ipm right now. I intend to replace the stepper with a servo since I have a box full of servo motors.

It's hard to judge movement on a compressed video. The motion compression algorithm is adaptive and when something suddenly takes off in a new direction it takes up to 15 or 20 frames depending on the codec to catch up completely with the action. This is especially so when the video has been optimized for minimum size as these have. The Divx codec has the option of optimizing for fast action as do some others but that always results in larger files.

[edit]

The amount of imbalance is very small, not enough to really be concerned about but I am fussy about that sort of thing. I don't think we have anybody that rebuilds turbos here. The mechanics here are doing well to change oil on a good day.

I have some more pics to post and will do so in a short while.

Evan
01-30-2008, 09:58 PM
Here is one of the other high speed spindles I built. It works ok but is limited to light engraving only. It still needs some work. I used the motor from a 30 year old Dremel tool that I never used all that much and was still in good order complete with metal frame motor. The business end has a pair of metal shielded (zz) bearings back to back that stabilize the spindle and take the side loads. Cooling air is sucked up the side of the housing at the corners and blown out the holes at the bottom. It turns about 30,000 rpm. The problem is it still has end float in the motor shaft. I need to figure out a way to preload it sufficiently from the top end. Not a big deal and it will work well when I am done. It chucks up in a 3/4" collet and is actually pretty rigid even with the long stick-out.

http://www3.telus.net/metalshopborealis/pics4/hss2a.jpg

http://www3.telus.net/metalshopborealis/pics4/hss2b.jpg

I also did a little engraving last night. I think it turned out well. :)

http://www3.telus.net/metalshopborealis/pics4/banshee5.jpg

lazlo
01-30-2008, 11:24 PM
Evan, I noticed that the CPU cooling fan you have on the spindle is crunched on the bottom -- did you leave something on the table the first time you ran the drill test? :)

lazlo
01-30-2008, 11:30 PM
Yes, it is. Yet Microsoft, in their infinite wisdom, doesn't support it because the codec is NIH, Not Invented Here.

That's not why Microsoft doesn't bundle an MPEG-4 codec.

The MPEG-4 license authority, in their infinite wisdom, decided that they would collect a royalty every time an MPEG-4 video is viewed, instead of licensing a codec to a vendor, like MPEG-2. The way the MPEG-4 license is written, they don't collect from the end user, they collect from the codec distributor. In other words, if Microsoft distributed the MPEG-4 codec with Media Player, everytime someone watched an MPEG-4 movie, Microsoft would have to pay a royalty to the MPEG License Authority.

What was particularly obscene is that the royalties were to be collected for multicast as well: if you broadcasted a concert over the web, for example, the codec provider would be billed for each person watching the video. Amazing.

So, MPEG-4 never took off commercially -- it was doomed to porn and illegally downloaded movies ;)

http://www.microsoft.com/windows/windowsmedia/licensing/mpeg4faq.aspx

Edit: this is also why Media Player doesn't support DVD playback without a separate, purchased plug-in to Media Player: the DVD Decoder Pack. The DVD Association collects a one-time royalty from each codec vendor, so Microsoft passes the royalty on to the end-use in the form of a purchased DVD enabling plug-in.

This is also why you have to purchase the DVD enabler (remote control) for the Xbox 360/720 and Playstation 2/3 separately -- you're actually buying the DVD Association license to play a DVD movie.

Evan
01-31-2008, 01:06 AM
That's not why Microsoft doesn't bundle an MPEG-4 codec.

Well, if it was invented by them they wouldn't have to pay a royalty, would they? :)

If you mean the dented fin on the heat sink, no, that was pre-existing. I think I dropped it. I drop a lot of things because my fingers don't work all that well. It's one of the reasons I have power feeds on everything and why I am interested in very lightweight materials.

BobWarfield
01-31-2008, 01:17 AM
Well, if it was invented by them they wouldn't have to pay a royalty, would they? :)

If you mean the dented fin on the heat sink, no, that was pre-existing. I think I dropped it. I drop a lot of things because my fingers don't work all that well. It's one of the reasons I have power feeds on everything and why I am interested in very lightweight materials.

MSFT didn't invent the MPEG stuff, I think Fraunhofer owns the lossy compression scheme and collects the royalties.

BTW, if you plumb compressed air, may as well use some for the spindle cooling too.

Best,

BW

PS We want more movies of high speed aluminum cutting mayhem!

Evan
01-31-2008, 03:50 AM
We want more movies of high speed aluminum cutting mayhem!

All you have to do is ask. :D

This one is cutting out a bearing locator plate 1/8" thick in one pass at about 20 ipm. Or maybe it was 30? I don't intend to do things this way all the time but I want to know where the limits are. This made a LOT of noise so I turned down the volume on the audio. The plate needs to be better supported in the middle to stop it vibrating. I was thinking of using a piece of sytrofoam insulation board or similar. I also need a mister system as you will see.

9.6 MB
http://www3.telus.net/metalshopborealis/banshee2.wmv

macona
01-31-2008, 04:11 AM
The Bijur misters are great. You can pick up a NOS tank unit off ebay for $80. Has a built in solenoid valve to just hook it to an extra output relay on you breakout board. Dosnt come with a nozzle but the guy gives plans on how to make them. Pretty easy.

I use Koolmist 78 in mine and it works nice.

http://cgi.ebay.com/BIJUR-Spraymist-coolant-unit-type-UB-Lathe-Mill-Grinder_W0QQitemZ360017187859QQihZ023QQcategoryZ63 3QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

S_J_H
01-31-2008, 09:52 AM
Evan, I am wondering why your machine is making pauses as it interpolates that cutout? Is that some sort of backlash comp?

When I do sheet cut outs like that I clamp the sheet firmly to the table with a piece of MDF as a backer and 2 sided heavy duty 3M tape between the sheet and mdf to prevent the part from getting nicked by the cutter when it is cut free. Was just doing a part like that last night in 1/2" 6061.

I use a fogbuster and it works extremely well. A bit pricy but it was a gift.

Steve

Evan
01-31-2008, 10:00 AM
That is Turbo CNC compensating for the accumulated math errors during an arc move. I wouldn't recommend TCNC for much of anything except for straight line cutting. It would be particularly unsuitable for a wood router as it would burn the work on curves like that. I should have Mach 3 hooked up today. I keep putting it off because I am having too much fun regardless of TCNC and it's bugs.

BobWarfield
01-31-2008, 02:38 PM
TCNC is not so good then. You'll love Mach 3 once its going for you. Those dwells are just silly.

I don't think the styrofoam would help you that much its so soft, but MDF would I bet. If you really needed it to hold still, looks like you machine two bolt holes. I'd clamp it solid to the MDF, machine the bolt holes, and then bolt through for more support in the middle.

For a lot of CNC jobs where you need access to the edges, they go in two steps. Machine through bolt holes with the workpiece clamped atop a tooling plate with a regular grid of threaded holes (easy for you to make!). Now install bolts to the tooling plate, remove the clamps, and have at it.

Enjoyed the vid. It gave a better sense of scale when this giant hand suddenly came into the frame with a spray can, LOL.

BTW, you could also consider a fogbuster-style cooler too with the Bijur or other coolant delivery. Plan data on my projects page:

http://www.cnccookbook.com/CCWishListFutures.htm

BTW, some stinker snagged that Bijur. Sorry!

Best,

BW

macona
01-31-2008, 02:56 PM
The Bijur is a fogbuster type of system. Just been around for a long time. I run at about 20 psi air.

ptjw7uk
01-31-2008, 03:20 PM
You could try double sided sticky tape to fix down to a sacrificial backing or even a vacuum table although I presume you would have to have some form of sacrificial media between the table and job but not sure how it all will work as I can see it would produce as many problems as it would solve.
Peter

bob_s
01-31-2008, 03:50 PM
Hmmm, CNC vacuum table.

Then make up sacrificial MDF vacuum surface on project basis

Evan
01-31-2008, 04:43 PM
A vacuum table is an attractive idea, at least for jobs with no side forces such as drilling. I already have a vacuum pump. The mister looks easy and I have all the parts including a solenoid valve. I don't do E-bay at all and don't intend to start. There are just too many possible problems that are doubly compounded when it has to cross the border.

I do use MDF but don't like to. It very abrasive and will take the edge off the HSS tools in short order.

bob_s
01-31-2008, 10:40 PM
Vacuum table attachment would be handy when making circuit boards- see that a lot with the CNC circuit board routers.

Engraving shouldn't have a lot of side loading.

Your high speed spindle would appear to be the perfect application for solid micro-grain carbide or ceramic tooling.

S_J_H
02-01-2008, 12:52 AM
I do use MDF but don't like to. It very abrasive and will take the edge off the HSS tools in short order.

Use some tape and you won't need to even cut into the MDF. Although I have never noticed any dulling on my cutters from MDF. But I am not milling it just using it as a sacrificial backer. So the tool might skim it now and then on the final pass.

Did this cutout tonight in 1/2" sheet with a 1/4" endmill. Notice the cutout portion is still fully seated and held tight by the tape. The cutter did not even touch the fiber board backer.

The fogbuster nozzle looks similar to that Bijur nozzle.
Steve
http://i109.photobucket.com/albums/n48/S_J_H/x3%20mill/toolingplate002.jpg
http://i109.photobucket.com/albums/n48/S_J_H/x3%20mill/toolingplate003.jpg

macona
02-01-2008, 01:28 PM
A vacuum table is an attractive idea, at least for jobs with no side forces such as drilling. I already have a vacuum pump.

Even side pressure is OK with a vacuum table. Commercial routers use this method all the time and the do all sorts of different sizes.

There are a couple companies that offer small vacuum tables for milling machines. I have seen them pop up from time to time. I have a Busch 2HP vacuum pump I have set aside for doing this.

I used carpet tape to hold down that chunk of koa for that barge. had to pry like heck to get it off.

lazlo
02-01-2008, 01:54 PM
The fogbuster nozzle looks similar to that Bijur nozzle.

They're actually quite different, although they achieve the same thing: they don't completely atomize the coolant.

I have both nozzles (the Bijur SprayMist from the same Ebay Volvo Dealer that Macona bought his from), and I built a FogBuster nozzle based on Karl Towsend's post on rec.crafts.metalworking.

The big difference is that the Bijur has a co-axial nozzle: the coolant is fed through an internal tube that's brought out to the very tip, and the compressed air flows separately around it, and the coolant/compressed air is not mixed until both leave the tip. This keeps the coolant from atomizing and generating a cloud of coolant fog that covers your shop and your lungs.

The Fogbuster combines the coolant and the air in the main body (the trapezoidal block in your picture), and uses a carefully designed monolithic (i.e., non-coaxial) nozzle to prevent the coolant from fogging. That's why the Fogbuster has a rigid nozzle -- if you deviate from the dimensions in the FogBuster design/patent, the coolant droplets will atomize, and you end up with a coolant fog like the cheap siphon-fed Kool-Mist or Porta-Mist units.

Evan
02-01-2008, 02:28 PM
I realized that I could safely use a very good and environmentally friendly coolant the other day when I was machining some Lexan.

Soapy water. My machine is nearly all aluminum and the few parts such as the ways that aren't are always covered in oil. Water works great on a lot of materials as long as it has some soap to lower the surface tension. Cheap too.

lazlo
02-01-2008, 03:18 PM
Sure, water with a surfactant (i.e., soapy water) is one of the best coolants :)

The nasty chemicals in commerical coolant are rust inhibitors and additives to prevent self-welding.

bob_s
02-01-2008, 03:45 PM
Soap = NaOH

I'd not recommend its use with predominantly aluminum machine tool.

NaOH is used to strip the protective aluminum oxide off aluminum parts prior to coating or painting. Worst spots for crevice corrosion would be around the grade 8 steel bolts you have used to hold it all together.

BobWarfield
02-01-2008, 05:34 PM
S_J_H, I love the tooling plates!

Your use of dowel pins reminded me to mention something. For those of you that want to learn more about this kind of fixturing, I highly recommend Carr-Lane's "Jig and Fixture Handbook."

It's not very expensive and there are endless ideas for setups in there.

Best,

BW

Evan
02-01-2008, 06:08 PM
Soap = NaOH

Not all the time, not even most of the time. Clear liquid dish soap/hand soaps are modified oxo or coco alcohols usually as the ethoxysulphate derivative. They are neutral pH and harmless to aluminum and your hands.

darryl
02-01-2008, 06:52 PM
Evan, a question or two regarding your high speed spindle. You made it from an R8 to MT2 adapter- I'm wondering how concentric the inner and outer surfaces are. If I bought this adapter to use similarly, am I apt to find some run-out that will p me off? Or are they pretty good, say within a tenth or so?

Another thing- did you make that dremel-type bit holder directly onto an MT2 stub arbor? I'm curious as to the method in your madness- what steps might you have had to take to ensure that your cutters run concentrically?

Another question- I can see where the area of the setscrew and the flat on a milling cutter would create an out-of-balance condition at that point- could there also be some effect from the drawbolt area? I'm wondering here if you have attributed any out-of-balance to what areas-

On the drawbolt- do you have a method of ejecting the MT2 from the spindle that doesn't rely on a smack with a hammer? On my mill I use a plastic hammer on the end of the drawbolt to persuade the arbor to come out, but I'd rather have some other way. I wouldn't want to be pounding on your spindle, especially the high speed one.

Evan
02-01-2008, 07:17 PM
In order;

I chucked up the adapter between centers so that whatever the location of the taper bore my bearing seats would be concentric to it. Since I machined off the original OD that concentricity is not relevant anymore.

The MT2 holder was a drill chuck arbor. I simply stuffed it into a 2 to 3 adapter in the spindle of the lathe and remachined the end to take spin saw collets. The adapters, collets and other tooling I have is all quality brands like Skoda, Bison, Dormer, Warrior, etc. They can be trusted.

It's possible that some imbalance is coming from the weldon shank cutters but mostly I have been using Dormer slot drills. They don't have a flat and are intended for collet use or a thread in holder.

The removal of the MT2 tooling is certainly an issue. However, the fact that I used an R-8 accessory for a spindle presents an elegant solution to bashing on the spindle. I remove the 5/16-18 drawbar for the MT2 tooling and drop in a 5/16 aluminum slug 1.5 inches long. Then I just screw in the drawbar from the R-8 spindle and it pushes the taper out gently.

bob_s
02-01-2008, 07:20 PM
PH neutral when dealing with sink full of dishes, soap concentrations are very low. Still at ph = 7.4 don't like to get that stuff in my eyes.

That all goes by the wayside when dealing with crevices as the concentration of the soap increases as the water evaporates.

Never use soap when washing my vehicles for the same reason.

Evan
02-01-2008, 07:43 PM
The generic name for the soaps used in most household liquid soaps is sodium lauryl sulphate. It's made from coconut oil. It's irritating to the eyes because it disrupts cell membranes, not because it is corrosive. PH 7.4 is only slightly above neutral to the alkaline side. That isn't enough or in the right direction to cause eye irritation. Our well water is pH 9.0 and it doesn't make your eyes sting.

darryl
02-01-2008, 09:50 PM
Thanks, Evan I see it now. The R8 adapter is internally threaded at the top for it's original drawbar. That gives you the opportunity to thread something in there to press out the taper. I should have realized that before.

I think I'm going to duplicate your idea for a spindle for my right-angle adapter for the mill. Hopefully the gear I plan to use will be machinable . If not, I'll duplicate it in epoxy and consider that a breakaway safety component. Coffee table epoxy is about the best I've found for this in terms of cost, availability, and toughness. I will also need the MT2 to MT3 adapter for my lathe, and some MT2 stub arbors. Idea duly filed away for this future project.