Driving the Knee? What do you think? and Why?

[Ron111]

There has been a good amount of interest lately with IJ's restored thread and chich's conversion of his knee mill. First off, why drive the knee, does it make a more ridgid machine and is it more accurate?

Is the basic asian 9X49 mill capable of being ridgid and accurate with the quill driven? What are the advantages and disadvantages?

Also when driving the knee, you also get an extra 8 ro 9 inches of programmable "Z" axis. With chich's conversion, an air cyclinder is used to assist movement of the knee where as it can be offset close to a balanced state where the servo is not lifting that much weight.

If the knee is driven and the servo is assisted by gas springs or an air cyclinder, will the Vertical hardened ways hold up? Getting the weight distrubution right could be tricky because of where the 150 lb table is positioned on the y-axis (front and to the rear) and of cource the positioning of the Xtravel where it is all the way to the left and to the right, this would place a lateral torque on the verticle knee ways.

I for one, would like to cnc an Asian 9X49, and drive the knee, but I need to address these type of dynamics concerning driving the knee. Also, are the Gib adjustments set up in order to support driving the knee. I looked at a couple of online manuels and only see an upper gib adjustment and in chich's conversion he mad a mod to the lower portion of the knee concerning adding a gib adjustment.

On the other hand, is there a problem with these mill as you approach the 5 inches quill limit with runout and are these mills ridgid with the quill extended in the 4 to 5 inches region, where they are able to take heafty cuts in the 30 to 40 ipm rate cutting 6061.

How much more ridgid are they with the quill locked and the knee driven. The machines which have converted and driven the knees, how are they performing and holding up?

Your thoughts,

Ron

[protman16]

Ron,

A lot of people here will tell you its better to drive the quill then the knee. Even IJ made a quill drive after he did the knee drive. Some will say that driving the knee is better because of greater travel.
I did a conversion on a 10x50 sized knee mill with the knee as the z axis. This machine has been running this way for about 2 years now. So far I cannot see any excess wear the slidway or have noticed any increase in slop. The knee is counter balanced so the servo does not need a lot of grut. It runs at about 135ipm max but I plan to increase that up to 500ipm soon. As far as positioning goes I can hold about +- 0.0002 in repeatability all day long.
The reason I went to a driven knee was that I thought the knee would hold its rigidity better over the full length of travel. So far so good.

Trevor

[turbostang]

when I was calling around asking about controls and whatnot I talked to the guys at Centroid. Their kits automate the quill on bridgeports. When asked why, the reply was the accuracy of the overseas copies are poor in the Z ways. When the movement was compounded over a certain amount of distance it would up bieng quite a bit at the ends of the Z travel. That is the only valid reason I could understand. I also brought up the deal about the quil hanging out 5" can't be good for accuracy. They claimed on a bridgeport it was not a problem in the slightest.... maybe they were just trying to sell their product, but I'm not sold on automating the quill. My overall impression was they are not big fans of most anything other than a bridgeport. I have the same machine as Chich, and I am still planning on automating the knee. I already have the air cylinder etc.
My machine came with some sort of certificate of accuracy, which may not be worth much more than the paper it's written on..but according to this paper it wasn't too bad. I think a disassembly and cleaning will do wonders for accuracy..because my machine was no where near clean when it was assembled.

[Ron111]

Trevor,
May be you could give me an idea as to the rigidity of the machine converted with the knee. The work that I would be using the mill for is cutting 6061 and any feel for feed rates on 6061 would be appreciated. I had talked to a vender of a square column type cnc system and he spoke of a lack of rigidity with the knee type of machines as far as taking healthy cuts such as .500 inch full with cuts at 40 ipm with .125 inch depths (4 flute carbide at say 4000 rpm). His thoughts were that the 9X49 were not ridgid enough for this task. I think that was his honest opinion, and I have never had opertunity to cut on anything but my shopmaster 3 in one. So, your thoughts about how much material removal rates that you would be comfortable with your machine would provide another calibration point.

Also, what kind of counterbalance system did you use on your 10 X 50. So far I have seen where the air cyclinders and gas springs have been used successfully. If you feel up to it please feel free to post some pictures of your build. We follow a good number of builds, but once there finished, rarely do we get to see how the machines are doing a couple of years later. All and all that is the true test of whether the engineering paid off, or what needed to be tweeked to get the bugs out.

Also where did you buy your 10 X 50? and any info about your configuration that you can share such which servo's (or steppers) driver's and control software (Mach 2)

Finally, it sounds like you are really pleased with your choice to drive the knee and 135 ipm is fast enough for production, but 500 ipm is flying, what type of changes will that take to diliver that type of speed to the knee.

Turbostang, pointed out that some info that was relayed to him concerned the possible trueness of the knee in some asian mills, and it's apparent that yours is dead on, that's where this thread may be an important point of calibration. IJ's thread is not complete and I could not determine whether on not the knee had a counterbalance system or what was the problem which did not allow the knee to be successfully driven, or if it was a manufactoring tolerence of the ways (or gibs).

Last off, some two years ago why did you choose to drive the knee (looks like the right choice for you), was it for the extra travel or increased ridgidity.

Again please feel free to post you build pictures or pictures of your machine, I need to see some good examples. Turbostang, same for you, I want to follow your build, it looks like a good machine!!!

Thanks

Ron

[turbostang]

I am getting ready to move so I may not be able to start on the coversion till the end of July. bummer.

The only thing I ask is What makes you want the 500 IPM? You can't really do anything to speak of in Z at that rate - but a couple hundred IPM would be good I'd think. Now with 20+ inches in Z, it would be a bummer to watch that thing rapid back and forth that slow

I have made several cuts on aluminum on my machine for the same reason you ask about. I had a 1" end mill cutting 1/4" deep with the power feed on a pretty good rate of feed and it handled it just fine and would easily take more. It had a hint of vibration but I think that was primarily from the sorry vise I have and the head wasn't trammed in.

I have big plans for my machine as well, but I can't mess with it for a couple of months.

[NC Cams]

QUESTION: why would you create a 500" per minute capable machine when it is all but impossible to bury a cutter into/thru most any material at that rate of travel? Yes, it makes for impressive jogs but, really, is it needed?

We have a factory built Bridgeport mill with a BMDC based servo controlled system. Essentially the same system was used on the mid-90's lathes, VMC's, mills and surface grinders that BPT made. Although it CAN run at huge jog speeds and rapid travels, we learned that higher accuracies were achieveable with much more "reasonable" feed rates. This is especially true any time you have direction reversals in the cutter path or when trying to do circular path milling.

When it comes to roughing, pretty much anything will work. However, when we do finish/smoothing/spring passes, we really slow things down - this pretty much eliminates follower error in the servo drives.

Although this may not be the case for steppers, none the less, the faster you try to move/do things, the more power has to be used/applied. Thus, a very stiff power supply will be essential for the rapid accelerations needed to achieve 500"/min infeeds. Although a 500"/minute machine will be impressive to watch, you still have inertia and mass to deal with. Don't stand in the way in any direction when something like that is operating.

Once you get it a huge mass of a milling head moving at 500"/min, something has to slow/stop it eventually and that will probably have to be the servo or stepper - I hardly think that way/guide friction should be enhanced to do the stopping as this will have to be overcome to get it to move that fast in the first place.

Yes, watching the machine jog end for end at a slow speed is as bad if not worse than watchine paint dry. HOwever, realistic expectations can result in acceptable performance without the cost penalties usually associated with designing, tuning and building lazer fast servo/stepper systems.

[Ron111]

NC Cams,
Thanks for you input.

But, what are you feelings about cnc'ing the knee on an Asian mill? Would there be advantage with the 10 X 54 with box ways(I think) over 9 X 49 with dove tails and if you followed IJ's thread (which is partial) what was the problem which resulted with cnc'ing quill as primary?

Add some meat to this post.

Ron

[Al_The_Man]

what was the problem which resulted with cnc'ing quill as primary?

IIR IJ found that the inertia/weight problem of operating the knee required a much larger force than he anticipated.
I have two Excello's that are similar to the B.P. in size and one has powered knee with air cyl. counter-balance system.
The quill is still the Z, personally I prefer it that way as in order to get the performance I have right now, would require a large change to the knee operator.
When you take the weight of the knee, the table, and if you have a part loaded that weighs 70~100lb and you want to peck drill! You can imagine the inertia of operating this load in high rate accel/decel in both directions.
If you power the knee and Z the quill, you can do a similar function I do when the knee is required to move down or up, which usually calls for a tool change also, in the case of my mill, I have no auto tool changer, just quick-change, so the M6 will be similar to a M0, I tool change lower the knee and press cycle start, the only proviso is with this process, you need to have an acurate display of knee movement, or a means of touching off again.
Al.

[NC Cams]

I am not in favor of moving the knee as the Z axis of a CNC mill.

The mass that you're trying to move is not in consequential. Even if you counter weight it, you're now DOUBLING the inertia of the system because BOTH the knee and the counterweight add up to inertia that has to be controlled. Eventually, friction comes into play and that too increases as you add more inertia.

It has been proven time and time again that reducing weight is better than increasing power when it comes to trying to make automobiles perform better. Since the mill and the auto follow the same laws of physics and inertia, the same logic should prevail.

It is hard enough to drive a spindle quill precisely with a Z axis drive and these have low mass and friction. I don't see how adding mass and inertia and friction (there is a bunch in the knee), will make it easier than driving the quill.

I do hope that folks who have done Z axis motion with the knee will chime in. Maybe they have found the holy grail on how to do it. I still contend that controlling anything with less mass and/or inertia will be easier that something with more of the same.

[turbostang]

Some good info and theory in this thread.

I can certainly understand moving the knee in a peck cycle, like mentioned above, bieng a problem. However, if you aren't using a counter balance but an air cylinder instead...you aren't moving the weight of the counterweight AND you are only moving a small amount of the weight of the knee.

Cheech mentions that he can move his knee up or down with his finger tips with the assist of the air cylinder. I am more concerned with the rigidity of the quill when it hangs out 5".

[turbostang]

For the records, I just went out in the garage and put a test indicator in the mill vise, extended the quill out the full amount, grabbed the back of the quill, pulled it towards the indicator and got a rather easy .004" movement. Keep in mind that is just me pulling on the quill and not the forces of a heavy cut pushing on it. Next, I pulled the quill back in the head and extended it only 1".. repeated the test and only got about .0025 with the same amount of pulling effort. That alone would heavily sway my opinion on weather or not to automate the quill or the kneed.

[Al_The_Man]

However, if you aren't using a counter balance but an air cylinder instead...you aren't moving the weight of the counterweight AND you are only moving a small amount of the weight of the knee.

I am more concerned with the rigidity of the quill when it hangs out 5".

The effect of the counter weight is to support or keep the weight of the knee etc in equilibrium, the required force to move the supported weight at the accelerated rate is the job of the servo to overcome the inertia required, this would be similar/same to a horizontal X or Y axis inertia demands .
With a C.B. system, you can only set an estimated weight, as the load will alter according to the part load weight. Even in a system I have, which is a constant force counter-balanced air system.

I have never experienced a problem with the quill extension, these mills were originally designed for manual milling using the quill, so I assume they have been engineered to also suit the conditions demanded by CNC.
Al.