Hydraulic ram position / pressure control

[lfe]

These presses usually have two hydraulic pumps, one with high flow and low pressure to move the cylinder quickly and another with low flow and high pressure for the final push.

 

[drbitboy]

Here is that article.

 

[Peter Nachtwey]

I am in Panama right now, so I don't have access to all the good stuff.
I can't even access deltamotion.com. Weird.
When hydroforming you are doing the water or oil based equivalent of blow molding plastics.
Anyway, Delta has experience at doing several different kinds of hydro forming going back to the 1990s.
It is possibThen one uses an intensifier where the pump pushes on a relatively large piston. It rod is connected to smaller diameter cylinder and piston. The forces on the two pistons will be the same but the smaller system will be able to increase the pressure by the ratio of the surfaces areas.

You can't control the pressure. The load dictates the pressure required to move or hold a position.

To be stopped the load and applied must be equal but there is a difference between a mass type resistance and what happens when you try to pressurize a volume.

Visit our website and look for applications.
deltamotion.com
I will be back and be better able to answer questions.

 

[Peter Nachtwey]

I am in Panama right now, so I don't have access to all the good stuff.
I can't even access deltamotion.com. Weird.
When hydroforming you are doing the water or oil-based equivalent of blow molding plastics.
Anyway, Delta has experience at doing several different kinds of hydro forming going back to the 1990s.
It is possible when one uses an intensifier where the pump pushes on a relatively large piston. It rod is connected to smaller diameter cylinder and piston. The forces on the two pistons will be the same but the smaller system will be able to increase the pressure by the ratio of the surfaces areas.


To be stopped the load and applied must be equal but there is a difference between a mass type resistance and what happens when you try to pressurize a volume.

Visit our website and look for applications.
deltamotion.com
I will be back and be better able to answer questions.

We have a sales engineer in Edinburgh. He can give you a big hear start.
[email protected]

 

[drbitboy]

If the current system's offset is repeatable to that 1mm total tolerance, i.e. if the hydroforming-side pressure pushes the die, against the hydraulic-side pressure control, back e.g. 5mm ± 0.9mm*, then would it be acceptable to initially position the die 5mm below the target position, and then let the hydroforming process push it back to that final target position? In other words, move the goalposts and declare victory?

* ±0.9 mm allows for the target-5mm positioning to have a 0.1mm tolerance.

 

[dwoodlock]

I think this could be done using servo hydraulic control. If you implemented an alternating control system you can command an absolute position for the upper ram. Then when the upward pressure exceeds some value, it would transition into force control and then back to position.

I have done this in metal forming applications with success, and you can even implement a variable displacement pump instead of using 2 pumps. Allows for faster movement when needed, and higher precision and sustained high forces when speed isn’t important.

A string pot wouldn’t work for the system I’ve used, but the concept is the same.

 

[ndzied1]

Well, according to the OP he needs the ram to remain in a set position. One thing is for sure, you cannot control position and pressure at the same time. This is not because of any specific parts or controllers. It is pure physics. It cannot be done.

As soon as a controller would switch to pressure control, the ram is going to move. Then you’re trying to play a game with the pressure command to stay at a position. But if you really just need the thing to stay in place you might as well just stay in position control. All you need is a system pressure high enough so that when the position control works against the hysroforming it has enough oomph to stay in place.

 

[drbitboy]

Tuning the system for position control with little or no upward force from the hydroforming side at the start of making each part is going to be very different from tuning the system when the part is near completion.

Pressure, or more specifically the differential between the area-scaled pressures, is what moves the die. Hydraulic-side pessure control is not the primary loop, but it could be the secondary loop in a cascade control. Ultimately the control loop's/loops' final CV (Controlled Variable) will be a valve position.

Part of the problem is that no process variable's value is linear, or even monotonic, with that valve position; it is the rate of change of pressure that is near-linear with valve position e.g analogous to a tank level control. Another part of the problem is that the position value is not is not linear with hydraulic-side pressure; it is again the rate of change of position that is near-linear with that pressure.

So this is a double-integrating (over time) system: valve opening ≈ dPressure/dTime; Pressure = dPosition(PV)/dTime; valve opening ≈ d²PV/dTime². Also, that valve opening ≈ dPressure/dTime is non-linear, because it is the differential between the working volume hydraulic pressure downstream of the valve and the hydraulic supply pressure upstream of the valve that determines the volume flow rate and therefore the rate of pressure change (via bulk modulus formula).

So any position error input (PV) to a feedback control system that results in a change in valve opening has to wait for that valve change to work through that doubly-integrated system to restore the position.

 

[MaxK]

Please excuse me for my inattention, but I can't see information about cylinder operation pressure rate, and cylinder dimensions (oil)

I.e. on what basis did you conclude that during operating pressure changes the cylinder displacement will be more than 1 mm?

 

[rak_plc]

Thanks for the input everyone, I've definitely got some food for thought. I've made some alterations for now to move into position control during the hydroform to try and maintain the position. I've got some variables I can play with and see where it goes. I'm currently holding the ram on machined stops for now while I try and control the pushback - these then need to come out when I'm in control of the pushback. The fixed stops can't stay in - I've already asked.

 

[Manglemender]

As others have said, this is about position control.

Question: why not have full hydraulic pressure avaialble all the time? I may be wide of the mark here but I suspect that using only a PLC rather than a motion controller, you would end up with a very unstable system under no load conditions.

It may be helpful to take a look at the valve charcteristics for the flow and pressure valves. Often on a flow valve the first couple of % spool movement is about "pressure buildup" and this in turn is likely to be a % of the available pressure so: if you monitor the spool position of your flow valve then you should be able to use that to govern the avaialble hydraulic pressure required to maintain a stable position.

Adding feed forward for the required pressure would definately be an advantage. Again, look at the valve profile as it may well be non-linear and previosuily I have used and Y-X lookup table to get a PLC output required to generate the force needed.

To hold position to +-1mm you would need a decent resolution form your wire pot in terms of bits/mm and I didn't see any mention of this. Analogue input cards vary a great deal in conversion speed and I would suggest that you need 2mS or better. With that in mind, It make no sense to run your postion loop faster than you can update the feedback or you will resposnt to the same error more than once.

In terms of accuracy, bear in mind also that, at those kind of forces, everything becomes a big spring so everything will bend - potentially including the anchor point for ypour linear pot.

Sounds like a fun project. Good Luck.

Nick

 

[ndzied1]

Words can only do so much. I tried to sketch out what I think the system looks like from all the descriptions so far. Would be better to have the schematic from the machine builder but maybe there is some reluctance to give that out??//??

I could be totally off but at least its something the OP can reference and say yes/no to each part..

 

[504bloke]

Would be better to have the schematic from the machine builder

Concur - I said that above, need a hydraulic circuit diagram, without that its all just guess work........

 

[Peter Nachtwey]

There is a difference between a press like in Norm's pdf and hydroforming, In hydroforming a fluid is pressurized to expand the work piece until it expands thoroughly into a mold. There may be times when parts of the mold move to elongate parts of the metal.

 

[ndzied1]

My model (not to scale and unpainted as it is ) was not quite clear then. The orange line was supposed to represent the metal to get deformed (shown flat before deformation). The “water” from below is pumped up and expands the part pushing it into an upper curved die moved by the upper ram.

I have never worked on a hydroforming machine but most videos I have seen run the top die down to hard stops and hold it there. I suspect this machine is intended to be anutomatically adaptable by having the top die able to be held in different positions by the hydraulics driving the top ram cylinder. It would then have to maintain position working against the force generated by the hydroforming fluid. This could eliminate the need to change toiling for different height similar parts. But again- all guesses.

It’s all just my interpretation of the dribbles of information we have so far so sure I could be totally off base. But the way I read it the task the OP is faced with has nothing to do with the hydroforming side at all. It is to hold the top die in place as the forces of hydroforming work against it. And I picked simple valve symbols as well. A real machine might have logic cartridge style valves for rapid moves but who knows.