[SGP]

I just recently purchased a new HAAS SL20 and its giving me about 0.012 inch part dimensional increase from morning start up till machine warmed up.

Does anyone have experienced this?

[gar]

060409-0728 EST USA

SGP:

Do you keep your room temperature constant day and night?

Are you almost continuously loading the spindle motor at 100%, or rapidly going back and forth from minimum to maximum speed?

With our machine at constant room temperature, and not heavily loaded we may see 0.001 to 0.002 change over a 1.5 hour period after machine has not run for 16 hours. This is not using HAAS's warm up compensation.

Your 0.012 change seems excessive. How many hours before this stabilizes?

Experiments you could run to identify partially the source or sources of the movement are:

These experiments need to be done on separate days after the machine has been off for a long time ( at least 8 hours ) and room temperature has been held constant before and during the experiment.

1. Do not run spindle. Mount an indicator from the spindle. Put a suitable item in the turret, maybe a tool holder or just a piece of bar stock, to provide a surface that is parallel to the spindle axis that will contact the indicator tip.

Write a program that will fast feed from machine zero to 0.1" from the indicator, slow feed to indicator zero, dwell for time to read indicator, rapid back to machine zero, and repeat. Select a fast feed and/or in combination with the dwell that will not overheat the servo motor.

Maybe only do the 0.1 travel to the indicator every X cycles to minimize exercising the indicator. X might be anywhere from 10 to 1000. Record indicator readings vs time. I expect somewhat of an exponential curve.

2, On a different day when the machine has again been allowed to thermally stabilize do the following:

Put a large diameter block in the chuck and turn a diameter for the an indicator to contact. This will be concentric to the spindle axis of rotation so we do not care where the spindle stops.

Mount an indicator on the turret to contact the turned gaging surface,

Write a program to position the indicator to read the gaging surface, dwell to read indicator, retract indicator 0.05", run spindle to maximum speed, dwell, stop, bring indicator to gage position, then repeat. Record indicator reading and time at each stop point. Again I expect an approximately exponential curve.

Again choices have to be selected to prevent overheating components.


3. Same as 2, but within machine limits, rapidly accelerate and decelerate spindle. Choose appropriate time points to measure with the indicator.

Use slow feed rate when moving turret with indicator.

.

[Geof]

That change is definitely excessive no matter how you are running the machine. Is the machine still under warranty and have you contacted Haas? The 1 to 2 thou change Gar mentions is what we get on our SL10s and HL1 and this is on bar work where the machine does not have any cool down time during loading and unloading.

[HuFlungDung]

Is the headstock bolted to the bed, or is it just sittin' there?

Maybe check that the X axis encoder couplings are tight.

[SGP]

I have contacted HAAS in which they came out and did some setting changes to the thermal compensation settings.

Now the machines is shifting about .004 inch from morning to night. This is after I run the warm up program.

However, I think this electronic thermal compension is not predictable. I still have to contantly check my parts because my part dimensions would jump about .003 before the thermal comp. kicks in and drops it back down to zero. Then this would continue through out the day.

I don't have a temp. control environment but I'm in Southern California where the temp. varies about 10 -15 degrees.

Also, I'm running my spindle load at about 40%.

Is this typical of a HAAS lathe or is this typical of lathes in general?

[gar]

060409-1614 EST USA

SGP:

At this time of year when we leave the furnace at the same setting thruout the day the effective ambient to our machine is probably within 1 deg or better because of the large thermal time constant of the machine.

This does not mean there are not thermal gradients within the machine, and there are. We do not use the HAAS temperature compensation, rather we adjust periodically. We do not have any step function changes.

The experiments I described I have not run but I should to get an idea of the characteristics of the machine. Just now the machine is making load cell transducers and I can not play with it.

Even though you have substantial temperature changes you might run the experiments I mentioned and get some idea of where the problem is or if it is distributed. If you run production without HAAS compensation you could track the change and the information might be sufficiently useful to develop a compensation macro.

.

[SGP]

When the HAAS tech. came out and changed my thermal comp. settings, I did a little experiment of my own where I checked and recorded part dimensions every 15 min. (kind of like doing an informal SPC). This is where I found out that the parts increased about .0005 inch every half hour , and maxed out at about .012 inch. So in order to get my parts to stay within acceptable dimensions, I'm constantly changing the offset every half hour.

gar: I will try your recommendation.

This is my first HAAS lathe so I don't know much about it. I bought it because it matched my budget. I do have a used 1998 Hwacheon lathe which does not behave as such. So before I make any kind of decisions in keeping it or selling it, I want to get some inputs from HAAS users (or even non HAAS users).

[gar]

060409-2013 EST USA

SGP:

We are very happy with HAAS and certainly recommend their product if you understand their design and limitations.

There is no internal cooling of a HAAS machine so you will get thermal changes. In contrast some other manufacturers cool their lead screws and other parts, but at higher cost.

To some extent you would expect the major castings to self temperature compensate because they all have about the same temperature coeficient of expansion. In the spindle and spindle motor area there is a lot of heat so castings in these areas will have differential expansion. Lead screws are a prime source for differential expansion.

Your 0.001" change per hour I would expect to progressively diminish as you approach 0.012" unless this is not the limit of change, or it is a result of ambient change. Do you just run 12 hours per day? When you start up the next day do you go back to the same settings of the previous day at that day's start up?

Without HAAS compensation you need to know what the change is every 1/2 hour thru the entire run time and what room ambient temperature is at each of these times. May be better what is the machine base temperature.

.

[SGP]

I run about 15 hours a day. And yes, when I start up the next day, I have to go back to the same settings of the previous day at that day's start up.

Is what I'm experiencing with my HAAS lathe typical or have I gotten a "lemon". I personally believe the best I can get out of my machine is .004" thermal change with the electronic thermal comp. settings on.

gar: Thanks for your inputs.

[PBMW]

I have not personaly seen a lathe that would not hold tol within .012. It is, in my experience, not typical of Haas lathes...
My shop undergoes pretty large temp changes nad I have none of those problems. I have an Okuma that will move a couple tenths in an hour, but nothing like what you are experiencing
I would want more of an answer than you've gotten from Haas.
I'm assuming you've only spoken with the rep and not the factory?
I have had MUCH better results from the factory than I have from our local rep

[gar]

060410-0902 EST USA

SGP:

The coefficient of thermal expansion of iron (steel) is in the ball park of 10/1,000,000 per deg C.

Assume a 90 deg F ( 50 deg C ) differential temperature, then per inch you would have 0.000,5 change. So you need large temperature differences and long path lengths to get anywhere near 0.006 to 0.012 changes.

You have not indicated whether the 0.012 is a diameter or radius, or the Z axis, the diameter you are working at, and the sign of the change.

Assuming you are talking about a 0.012 diameter change, then that is a 0.006 radial change and the expansion figure we are concerned with.

The X-axis lead screw reference is at the top rear of the machine, I believe. Expansion of the lead screw would therefore make your diameter get smaller.

Expansion of the spindle casting will be vertical and this will work on the cos effect of tool position relative to part center line. This should be a small factor.

The HAAS compensation function only has a small range. I am not where the manual is, but my recollection is 0.002". Whether this is radius or diameter I do not know.

RC (resistance - capacitance) voltage curves and thermal resistance mass temperature curves follow an exponential curve. So in one time constant you change 63%, in the second time constant you move 63% of the remainder, and so on. This is why I expect your correction to change with each new equal increment of time.

That your machine returns to the same starting point each day means there is nothing taking a permanent offset, such as a loosely mounted component. It really seems to imply something is getting quite hot, or there is something else that can affect the position system.

It is quite unlikely that you are talking about the Z-axis since that should self-correct each new part.

If you can disable the HAAS compensation and get a good curve of the variation per equal unit of time that can maybe tell us more.

.

[gar]

060410-1025 EST USA

SGP:

Is there any possibility that the HAAS compensation needs the sign reversed?

If the sign was wrong, then you would compound the actual machine thermal change.

[miljnor]

This is definitely machine and possibly user dependent. I have all HAAS machines and we don't have temperature control in our shop. I get about a .001-.002 varience for temp warm up and after warm up it will stay thier all day unless the day is really cold. Then you have to warm up again after lunch.

I bought one of my lathes from a guy down the street from me that was having similar problems to yours. But after buying it (got it for a steal! ) it worked great for me, so I cant help thinking that it was a programing/user problem.

.012" is realy excessive and I would definitely keep on haas for a fix. the .004 they got it down to is better but still rather large for these machines.

[100]

This has to be a control related issue. Something is wrong with the control or a setting in the control.

[gar]

060413-0921 EST USA

SGP:

Any new information on your SL20 problem?

.

[SGP]

Still working with HFO regarding this issue. HFO technician changed some more parameters. However, still seeing about .004" tolerance shift.

[werks]

Sounds like you are on the right track.

I have owned several Haas lathes, and have never seen more than a couple of thousands difference.

[gar]

060413-1637 EST USA

SGP:

Playing with parameters without knowing the root cause of the variation seems to be the hardest way to correct the problem.

As several have mentioned 0.012" variation is way out of line with expectations.

We need to know if this is a diameter or radius change. I am still assuming it is the X-axis that you are worried about, and your change is diameter. This means an X-axis physical change of 0.006".

See Temp. change sl20 http://www.cnczone.com/forums/showth...e+compensation

.

[gar]

060415-0740 EST USA

SGP:

On our SL20 on the X-axis the distance from machine home to the turret pocket centered on the axis spindle is about 11.5/2 = 5.75". This may be approximately the length of lead screw that thermally might affect tool position.

Using my thermal expansion calculations from post #11 above we get 0.000,5 * 5.75 = 0.0029" change for a temperature rise of 90 deg F. This assumes the entire 5.75" length is raised 90 deg.

A 90 deg rise above 70 deg is 160 deg for the temperature of the screw. This is quite hot and furthermore the entire 5.75" would not be uniformly heated to this temperature. Thus, other areas would have to be even hotter.

Someting just does not make sense here.

When I ran experiments on the HAAS compensation algorithm that I discussed in post #11 of (See Temp. change sl20)
http://www.cnczone.com/forums/showth...e+compensation
I was not able to find a reference to a variable that contained the compensating value. This meant that I could not watch such a variable and/or a derivative, and see what was happening. Further, HAAS apparently totally hides this compensating value and you have no way of observing its effect other than by an external means of measuring the X-axis position. Quite unsatisfactory from a troubleshooting point of view.

.

[GAS]

SGP:
Your thermal growth problem is easly fixed. Why don't give the factory a call so it can be properly attended to. My SL-20 after adjusting the thermal growth paramaters will hold .0005 all day long with out an adjustment. Rum the spindle for 30 min, them make parts. If you go to lunch just let it set. If you run the spindle thinking that the machine will change size it will be all messed up. I found this out by doing just that. I had to chase the offsets untill it settled down. Then the next day I let it set went to lunch came back and it was right where I left it. 1018 steel at 800 sfm .020 finish cut 1.500 dia +.0005 -.0000 the bore was with in .0003. Haas cares about all of there customers and if the dealer did not do a good job, then the factory will fix it today!