Hydraulic Straightening Press help 1

[kquirk]

Hello,

My company manufactures ballscrews, one of the most time consuming parts of our manufacturing process is straightening. We use a hydraulic press to straighten the steel rods of varying diameters. I have been tasked with creating a chart or table that will tell the machinist how much deflection he needs to get to reach our tolerance for every shaft. My problem is that many of the ideal deflections to straighten the parts out are in the elastic region of the stress strain curve. Currently does anyone know of an equation to use to help me? I am going to have my machinist keep track of data from now on like the initial TIR and deflection put on the shaft as well as the final TIR. I know there are companies that specialize in this field, I have searched some of their websites but have not found any real help there.

Thanks,

 

[berkshire]

kquirk
I think you are going to have to do this by empirical means.
Part of the problem, is that you have to deflect the bar to the limit of the elastic curve first.
Then, and only then, can you add additional deflection to straighten or curve the part.
Different metals have different rates of spring back, which are well known and are available in chart form, you can use this for a first cut.
However you will also have in house variables. such as spacing of support blocks, existing radii in the bar, variations in temper etc..
Your machinists will be able to log this information which will help, but here is no cookie cutter approach to this.
B.E.

You are judged not by what you know, but by what you can do.

 

[dhengr]

Kquirk:
You might be better off mechanizing and working to quicken the process so that a pressure and deflection can be applied, and the deflection measured in process; then released and straightness measured again; then the pressure ramped up as a function of the change in straightness. The point being that you want to hit the bar 4 or 5 times in a minute and measure/view the straightness to dictate the next step’s pressure. Maybe a laser line btwn. the inner surface of the two support blocks (that’s the inner edge, concave edge of the bar along its length axis) and when the bars stays on the line when the pressure is released, you’re done. Note that if you do this straightening before machining the ball races, you will likely get some relaxation or return to the earlier shape. If you do the straightening after machining, it might stay better, because you are not releasing the residual stresses during machining.

You should be able to hone in on the problem by making some calcs. for the range of force/pressure vs. distance btwn. support blocks (span length), and vs. shaft dia. (section properties) and material properties. You know you have to strain the bar beyond yield, so that’s a starting point. And, over time you may learn something about increasing the starting point as a function of out-of-straightness for a given dia. bar. There may be some advantage in setting/changing the support block spacing as a function of bar dia. for a more constant pressure/deflection curve. I assume your support blocks and pressure block are shaped to match different dia. bars, and they may also be radiused to apply their force over some length during bending. There may also be some advantage to having only a few people do this process, not every machinist in the shop doing his own. I have watched and worked with people who were incredibly good at heat straightening and force straightening (we called the machine/press a bull dozer). Ask them how they did it and it was all gut, intuition and experience, how it acted after the first couple bumps, told them what, and how much, to do next. I could calc. some of this action and come pretty close in terms of heat input or force and the locations of application, but they just did it all by feel and by eye, and darn well, I might add. Of course, we were working on bigger members or weldments and not usually working to thousandths.

 

[chicopee]

I am not sure if strain stress curves would help but at the least, I would keep a catalog of such curves. Personally what I have done to figure out how much strain parts that were formed with dies was to get blank pieces and scribe mesh patterns either in rectangular or polar forms, of known dimensions before these blanks were cold formed. The above responder stated "empirical means", that's the approach that I would used along with scribbing the intact pieces with mesh patterns of known dimensions to calculate strains along with recording the forces imposed on the test specimen in order to develop graphs or formulae.

 

[Compositepro]

I don't think that calculations will help you at all. You need a dial indicator mounted under the shaft during the straightening process so you can see how much the press deflected the shaft and then to immediately see whether there was any permanent deflection after the press is released. The shaft should be in V-blocks so you can also rotate it as needed to check run-out. The process should be controlled by measuring strain and not pressure.
There will be stored stress in the shaft as a result, and machining afterwards will affect straightness.

 

[racookpe1978]

You CANNOT think about theoretical solutions or curves or anything else until you KNOW as close as possible what is going wrong!
How many rods are bent (need straightening) compared to how many are built per day/week/month/year?
What is your spec for straightness, and how did you determine it?
Can that spec be loosened - or is it fixed by the process you are using the part (the rod) for?
Do you know why your rods are bending, and what can you do to eliminate that problem - rather than waste time (MONEY!!!!) trying to fix something after it has occurred.
What diameter rod?
What length or lengths of rods?
Where (along that length is the "typical" bend point - the place where the deflection is) is is that deflection even along the length of the freely held rod or "kinked" at a single point?
A single kink or gradual bend or several?
If several, are the bends (kinks?) all in one plane, or are they in different directions?

 

[Tmoose]

Curious things can happen when straightening parts with built in stresses, and especially when straightening my own over-correction.
Bauschinger effect.

 

[berkshire]

Moose,
I do not think that the OP will have much trouble with the Bauschinger effect. Unless one of his operators totally messes up the straightening operation.
If that's the case I guess you can call that free work hardening/ softening.
B.E.

You are judged not by what you know, but by what you can do.

 

[kquirk]

Thanks for all of the advice,

We do use a dial indicator to measure the deflection. We currently straighten the shafts before and after machining. The diameters vary from 8-50 mm and the lengths vary from 500mm to 4m. I have my machinist logging the info now, I plan on creating the table from his information.

Thanks again for all of the responses, I had a feeling that we needed to do it this way I just wanted to throw the idea around. I made some calculations and proved why my machinist is having a hard time with the smaller diameter shafts and not the larger. The 8mm shaft needs 0.14mm deflection to yield and the 50mm shaft needs a deflection of 0.02mm to yield.

 

[Tmoose]

Hi Berkshire,

"one of his operators totally messes up the straightening operation."

Have you ever personally straightened any precision spindle shafts, steel bars, crankshafts, crash or drunken wheely bent motorcycle fork tubes, etc?

regards,

Dan T

 

[berkshire]

Yes, I used to do it for a living about 15 years ago.
B.E.

You are judged not by what you know, but by what you can do.

 

[berkshire]

To elaborate,
Chrome moly spindle shafts, Aluminum aircraft components after heat treating, and titanium components after chem etching.
Bent motor cycle tubes no, aircraft tubes yes.
B.E.

You are judged not by what you know, but by what you can do.

 

[tbuelna]

This is one of those situations where a skilled technician using manually applied force (through something like a lever or jackscrew mechanism) could possibly produce a better result in less time than could a worker with lesser skills simply inputting commands to a press machine applying hydraulic force. It is quite difficult to accurately control force and displacement using a simple hydraulic cylinder. A skilled worker manually apply force through a lever or jackscrew would allow far greater control in where and how the straightening force is applied because there is a more direct/immediate type of feedback to the operator.

A good example of the high level of accuracy that can be obtained in a manual operation is the process used to straighten rifle barrels. There are no modern machines that can come close to matching the speed and precision of a skilled technician using only his eyes, hands and a manual screw press to straighten rifle barrels.

Something to consider.

 

[berkshire]

tbuelna,
I would heartily second that, most of the smaller parts I used to hand straighten were done on an arbor press, using tactile feedback.
When straightening precision shafts, we would use a set of balancing rollers, set the shaft to be straightened on them and give it a spin.
The shaft would stop with the bow down, we then knew where to straighten next, When the shaft would stop at any random position ,it was considered straight.
A double check with a dial indicator would usually show it being less than .001" out of true.
Only when the parts were too big to hand straighten, did we go to the hydraulic press.
B.E.

You are judged not by what you know, but by what you can do.

 

[PNachtwey]

It is quite difficult to accurately control force and displacement using a simple hydraulic cylinder

You can't control force AND displacement at the same time. The position or force can be controlled OR the force or position can be limits such that the rod is bent to a certain position not to exceed a force limit or a to a force not to exceed a position limit.
As far as accuracy, positions can be measured very accurately with MDT rods and if that is not good enough then use glass scales. Applied force can be controlled very accurately, much better than a human looking at a gage. If precise force control is required then use a load cell.

On top of that. The hydraulic system can measure the force as the rod is turned. If the forces vary the rod is not straight. Even a small offset in position will cause the force to vary a lot. The goal would be to turn the rods or shafts and detect no force applied as the rod is turned.






Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

 

[tbuelna]

PNachtwey-

In practice, there are some issues with accurate control of small displacements in most hydraulic press systems. First of all, the hydraulic circuit feeding the cylinder normally operates at a fixed pressure, which means a constant fluid pressure force on the rod/piston. But there can be large instantaneous changes in the opposing force on the cylinder, especially where there are small incremental displacements requiring stop/start movements, due to static friction effects caused by the cylinder seals, bushings, etc. Unless the cylinder piston/rod area is extremely large and/or the circuit pressures are extremely low, it would be very difficult to achieve the high level of linear displacement control (ie .02mm) described in the post.

In the picture shown in my previous post of rifle barrels being manually straightened, what you can't see is the optical target on the wall that they are aiming at which generates a reflected interference pattern within the barrel bore that provides a very accurate indication of misalignment.

 

[PNachtwey]

In practice, there are some issues with accurate control of small displacements in most hydraulic press systems.......

Catch up. There have been a lot of advances in hydraulic position and force control over the last 30 years. You probably haven't met anybody knowledgeable in the field yet. Precision isn't a problem any more given the right design. The immediate problem would be where and how much force to apply to straighten the pipe or rod.

If a load cell could measure how the force changes as the pipe rotates I think it wouldn't take too long to figure out the programming.

Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

 

[Tmoose]

If the ball screws are particularly hard or otherwise lack ductility then straightening by bending may bring risk of cracking.

Tufftrided or carburized crankshafts and iron castings are often straightened by inducing compressive stress on the "low side" by peening with a blunt chisel.
I was looking for an old picture I'd seen of an OEM cam shaft blank straightening department armed with air chisels.
This is as close as I could come.

http://2.bp.blogspot.com/-zjTtQRAkpMs/UUl_P-dYSEI/AAAAAAAABB4/6G2sNWGZ-78/s1600/DR-1105-CAM-03.jpg