velocity - hydraulic motion

[TConnolly]

If I understand you the piston operates a lever which moves the cart 1 meter. Assuming we are talking metric tonnes here and not USCU tons, the cart has a mass of ~150 kg, but because of the mechanical disadvantage of the lever the piston sees a load of 250 kg and the piston stroke is less than 1 meter. Is that correct? If so then what is the distance the piston travels? Once we know that we can calculate the acceleration and required force.

Based on the info so far this beyond the capabilities of most PLCs for motion control. But that doesn't mean we can't help - some of the forum contributors are motion control experts.

A sketch will be really helpful.
Other pieces of helpful information include the hydraulic operating pressure, pump size if it has already been selected (hopefully it hasn't as it might be inadequate), and cylinder size if it has already been selected (again hopefully it hasn't as it may surprise how big it needs to be).

 

[leitmotif]

1 meter in 0.6 seconds equates to
1.6 meter per second
100 meter per minute
and 6000 meter per hour or 6 km/hr
These are all based on average velocity and do not allow for accel or decel.

Now then to get rough with numbers and rounding off
that equates to 3 feet per second and 500 lbs. Not allowing for rolling friction that is approx 3 HP. If rolling resistance (friction) is 0.1 then that is only .3 HP (granted at steady speed - so add .5 HP to give enough beef for acceleration and maybe more important decel)

I would give some consideration to using VFD and a suitable size motor. Even if this is constantly reversing at a cycle time of 0.6 second (one direction) or 1.2 second for "out and back", with adequate cooling on motor and VFD it may be feasible. VFD will have ramping for accel and decel which can be adjusted to get desired results.

What is wheel size and what are you running on - steel rail or?? And I assume this is all flat (give or take 0.25 cm)

What is time factor for loading the cart and for unloading? This would allow for motor cooling.

What is material on cart - if block of steel then mayb accel and decel are not important - if "cannonballs" then maybe you will really have to watch accel and decel.

Figure 100 to 150 for motor
150 for gearbox
150 for VFD
150 for chain and sprockets

Hydraulic pump motor and cylinder are probably as much and you still do not have controls nor the lever and all the other stuff to make it work.

Dan Bentler

 

[TConnolly]

From the posted graph it appears that Hecttorr wants to accelerate for 250mm, run at speed for 500 mm, then decelerate over the last 250mm. To move the load following that profile in .6 seconds then the peak speed is not 1.6m/s, it is 2.5m/s.

To accelerate 150kg to 2.5m/s in 250mm, a=12.5m*s^-2, the required force applied by the lever to the cart is 1.9 kN (about 425 pounds force). That is what is needed to get the load moving. The actual force will be greater because we haven't yet accounted for friction or any other losses. To determine the piston requirements we need more information.

 

[Peter Nachtwey]

The formula is
MaxVel=1.5*Distance/Time
AvgAcc=4.5*Distance/Time^2
Trapezoid ramps shouldn't be used. Trapezoidal ramps assume the acceleration rate can be changed instantly which is not possible.

However, the average acceleration must be multiplied by about 1.5 to get the peak acceleration. The about 1.5 depends on the type of s-curve. S-curves should/must be used when accelerating and decelerating as fast as 12.5 m/s^2
To design the hydraulic system I have my Mathcad simulator that solve 8 differential equations but this one.
http://www.deltamotion.com/peter/Maxima/VCCM difeq.html
This version doesn't take into account the bulk modulus of oil but should be accurate to within 2% which is better than the variance of the part specifications. You must size the cylinder and valve so that the hydraulic motion profile is faster than the desired motion profile.

 

[Honga]

This seems like a student question.

If so please instruct. We can help you better if we know these things!

 

[hecttorr]

Seems more complex than I knew.

The instalations hydraulic and electrical are realised already and I need to made the PLC software. I can not use anymore motion controller, I must made this in the PLC software.

I must write some simple motion ecuations in PLC yet, if complex motion can not write in PLC.

I have attached a sketch with lever from cylinder.

 

[ndzied1]

Please tell us what type of positional tolerance you need to hold at the end positions.

 

[hecttorr]

Please tell us what type of positional tolerance you need to hold at the end positions.

Tolerance of end position = -/+ 5mm

 

[hecttorr]

from Alaric : If so then what is the distance the piston travels?
The distance the piston travels = 400 mm;

hydraulic operating pressure = ~100 bar;
pump have 15 kW ;
yes, for this lever the piston need more force, bat give more speed from cart

 

[hecttorr]

from leitmotif: What is wheel size and what are you running on - steel rail or?? R: Yes, cart running on steel rail.
For "cannonballs" I need to accelerate and decelerate.

 

[Honga]

Can you start from a rest state, increase the output from the pump, measuring hydraulic pressure, until the piston starts to move?

The value of pressure*SA of the piston should be a (For sake of simplicity) a constant for overcoming static friction.

Although the above will vary with many parameters ie: wear of the O-rings in the piston, introduction of heat, wear on the wheels or rails/ dirt on anything.

All of the above indicate to me that it would be nearly impossible to get an accuracy of +-0.5% with straight feed-forward techniques. Even though you stated you can measure position/velocity with your encoder you have made no mention of algorithmically feeding this value back.

You should do a google on Peter N's suggestion for an s-curve, this will eliminate jerk, and increase lifetime of everything.

 

[ndzied1]

I must made this in the PLC software.

And if this is physically impossible then what?

 

[Peter Nachtwey]

I know! I know!

And if this is physically impossible then what?

Call Tom Jenkins!!!!

 

[leitmotif]

I have attached a sketch with lever from cylinder.

OK now that you have the mechanical side of system installed:
1. Have you tested for operation
2. As installed is it able to do the acceleration then "move" then decel as you desire. Yes I know this would take manual operation of the valves (I think you should ALWAYS be able to manually operate them) and you would have to be pretty quick at that
BUT can you get close to show it can work as desired?

Think I would use a one meter cylinder to get the cart to move one meter and throw out the lever.
Dan Bentler

 

[milldrone]

hecttorr and other forum members,

I browsed this thread twice and did not see if the move is the same distance every time. If the moves are the same each time point to point, then a crank with a 1 meter throw might fit the bill nicely instead of a fluid power cylinder. Especially with the "no motion controller requirement".