Control of 6 to 25 motor through VFDs for lifting load 3

[Skyba]

Hi everyone,
I am a student working for a company and I have an interesting task. The project consists to lift different heavy loads (15 to 50 tons) with the help of 6 to 25 motors. The amount of motors depends on mass and geometry of the load, and it should be easy for any operator to add/remove a motor.

I have outlined a plan for the electric configuration :

http://img404.imageshack.us/img404/7163/ohzy.jpg

- Required individual motor power is less than 1/2 HP (200-300 watt per motor)
- Motors are tri-phase asynchronous. All motors are the same model.
I am using one VFD for 5 motors because :
- it seems to be the cheaper solution (VFD can control up to 5 motors according to constructor)
- i do not require exact speed/torque regulation (lifting time is 15 min, over 2 m)

However,
Since the motors are working together to lift the load through chain hoists, and the load needs to be lifted on a horizontal plane, it is important that one motor does not lift faster than the others. I have read that the charge would distribute naturally because of "slipping" of the motors, given a high slipping frequency.
If not, the inclinometer would jump in. If it detects an uneven lifting, the PLC would stop the system, and the operator would level back the motors by opening switches on the motors that are "ahead". This should only be an exceptional situation.

And now with my question : Is my outlined draft a good (and affordable) design for sharing the load between the motors ? Should I interconnect my VFD's ? Do I even need inclinometer and PLC ? I am quite new at this and I am shurely missing some points here.

Thank you in advance for your answers ! Every help is greatly appreciated.

Basile

 

[itsmoked]

That looks like it should do pretty well. The slip will distribute the load as suggested. For something this big/expensive I'd include the extra monitoring.

Normally this would have brakes all over the place. If power fails what stops it all from crashing back down.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[LionelHutz]

I'm a little confused about the use of so many motors. Do you have a winch and cable for each of these motors?

I could see you having trouble with starting these lifts though. I could see you having difficulty with the winches all taking up the load if you're using a different number of winches for different sizes of loads. Since some winches have been off, they'll be starting will different lengths of cables so you may be requiring some motors/winches to slip at locked rotor conditions until the other motors/winches catch up.

With the system you're looking at, the next logical step would be to build the system so the inclinometers and PLC can control the speeds of the different VFDs to automatically level the load. Depending on the VFD, you would do torque control to automatically load all winches to ready for the lift and then speed control to keep the load level.

 

[Skyba]

Thank you for the kind responses.
The system functions with a chain hoist for each motor. It would help lift up tank rings of 8 to 15m diameter, as depicted on

http://www.in.all.biz/img/in/catalog/256202.jpeg

@ keith : The chain hoists have a mechanical locking which prevents the load from falling down if no torque is applied.
@ Lionel : The hoists are linked to the motor with a chain gear - unlike a winch, the chain hast no "start" or "end".

With the system you're looking at, the next logical step would be to build the system so the inclinometers and PLC can control the speeds of the different VFDs to automatically level the load.

That would require a VFD for each motor, right ? Is it worth the cost ? An closed loop should not be necessary if the mass has a good distribution over the motors. The inclinometer here is only used as a "unbalance detector" that would shut down all motors. The operator would then switch to manual mode and choose which motors to activate. The inclinometer may not be necessary though.

My hope is to get your opinion on the ability of this system to load a mass evenly without any torque / speen control. We would like to build an easy but safe system.
Thanks !
Basile

 

[SceneryDriver]

I'd vote for a small drive for each motor, as it would be the most flexible.

1/2HP drives are pretty cheap and you'd save the cost (and panel space) of the overloads and fusing as well, since the VFD-per-motor solution would handle that.



SceneryDriver

 

[Compositepro]

All the motors will run at close to, but not exactly, the same speed. If one motor is going faster it will pick-up more load and then slow down (good). However, only the inclinometers will keep the load level over time. The platform can slowly tilt and all the motors will still carry close to equal loads. Depending on how the lift is rigged, tilting could shift more load onto the slower motors (bad). For a single lifting event your system should work okay. You will need to be able adjust each motor individually to get out initial slack. Then you need to be able to adjust the groups to correct for tilt.

 

[Skyba]

Hi SceneryDriver,
Thanks for your post.
If i sum up, for each group of 5 motors we have :
In a 5 VFD configuration
5* $250 for a 2.2kW drive
= TOTAL $1250
Is it more flexible ? To get a good load distribution, we would need to connect the VFD to between each other to inter-regulate Torque and Speed, as suggested LionelHutz. The system overall would be more complex, woudn't it ?

In a 1 VFD configuration
5* $30 Thermal Overload Relay
5* $20 manual switch
$350 for a 2.2kW drive
= TOTAL $600
This seems to me the fastest setup. As showed on my initial drawing, we could just simply connect 5 motors to each VFD casing, making them easy to plug/unplug. However, I do not know how good the load share is if we have a 3 VFD - 15 motors situation, for example.

Does anyone have any experience in using that many motors to share a load ? If the system is reliable I might also get rid of the inclinometer/PLC...

 

[Skyba]

@ Compositepro : True, i haven't thought of the situation of a motor being too slow, catching up more load and thus not being able to catch up. But how could one correct that ? Let's suppose the tank slowly tilted towards one side and 3 motors out of 10 are too (s)low. To level back all the motors we some of them would be momentarily overloaded, because the configuration doesnt allow us to lower and raise the position of opposite motors simultaneously.
The only way to deal with this would be to hope that it doesn't happen or that the loads would carry close to equal loads anyway... Or to update the whole configuration.

Note : Today, the tanks are lifted manually, with one person pulling the chain hoist at each jack

 

[LionelHutz]

Are there people you can talk with that have experience operating this type of system in that application?

I would want 1 VFD per motor unless I can talk with people who have used similar systems and know from their feedback that manual intervention is a rare occurrance. I would also put enough PLC to VFD connections to allow individual control of each motor. Then I could try it first without any levelling controls first and this leaves the system flexible enough that I could implement closed loop level control if necessary.

It would be silly to have ~most~ of the hardware installed to eliminate the manual intervention and not do it if necessary. Just saying.

A ring like that might be able to be leveled with 5 motors on common VFD control if each VFD motor group is together in one area of the ring.

Here is something that hasn't been defined. Say you have all 5 motors lifting a part of the ring. How many are really needed to lift the load? I hope you can see where this question is leading. The load might not be equally supported unless you control each winch.

 

[jraef]

The amount of motors depends on mass and geometry of the load, and it should be easy for any operator to add/remove a motor.

I still don't understand this requirement. The geometry I can grasp, but changing the quantity if motors due to mass is unnecessary just use a motor big enough to handle the largest mass. Then you have one drive per motor based solely on geometry. Skewing of loads with multiple hoists is already tricky enough, you gain nothing by complicating it more by trying to do it across multiple small parallel motors.

Or perhaps this begs the question, what is it that you perceive you will gain by doing it that way?

"Will work for (the memory of) salami"

 

[Skyba]

I would want 1 VFD per motor unless I can talk with people who have used similar systems and know from their feedback that manual intervention is a rare occurrance.

Yes, I lean towards that too, given your comments. Unless someone on the forum knows for a fact that controlling 20 motors with 4 VFD doesn't need manual correction.

A ring like that might be able to be leveled with 5 motors on common VFD control if each VFD motor group is together in one area of the ring.

That's the case. But even then, I guess interconnecting the VFD's would't bring much, because every VFD is in V/f mode. Torque or speed control works only for 1 motor per VFD.

Say you have all 5 motors lifting a part of the ring. How many are really needed to lift the load?

That's a good question. If I buy motors that can barely lift their supposed share, then motors that are "ahead" would slow down by "slipping" under the excessive weight. But as poited out Compositepro, if motors are too slow, they would not be able to catch up.

just use a motor big enough to handle the largest mass.

If we do that, then some motors would be strong enough to lift a large part of some smaller tank's mass. The problem is that the hoists too are a limiting factor, because they have a maximum load of 5t. We could buy stronger hoists, but since we have a big stock of those here, the price wouldnt not be worth it. I am trying to give a 2.5t nominal load on each jack, with a maximum load of 5t to account for wind and tilt.

To get back to the initial problem, I get more and more the impression by reading your comments that, when not sure, one should lean towards the safest option. The system is pricier and complexer though and I wish someone could tell from experience which works or not. I will try to make some calculations over the maximal acceptable tilt angle so that each post does not carry more than 5t. Maybe it turns out that even a 1% speed difference in the motors would not make huge difference in the load distribution...

 

[iop995]

If mass is almost egualy distributed across ring, I think best way is to spread hoists across ring. For higher mass ring, multiple VFD with allocated hoists will be interspersed to increase lifting force. For example if a VFD feed 6 motor hoists at 60 degree apart each other, will be able to lift a ring with 15t, maxim 30t; if mass is higher add another VFD with 6 hoists interspersed between first hoists. In such arrangement will be avoided tilt problems.

 

[LionelHutz]

You have to be careful of the start conditions. Can one winch be first to take-up the load? A motor can draw 5 or 6 times its FLA rating if you slip it too much. It can also produce 2 to 3 times its rated torque as it slips past the breakdown torque peak. You will need to ensure the VFD can supply this current and the winch can handle this torque if this is a case that can occur.

 

[Skyba]

Before I answer, I'd like to re express my gratitude for everyone helping me here. You guys are really awesome !

@ iop995 :
Do you mean something like

http://img845.imageshack.us/img845/2486/yyra.png,

figure 1 ?

In such arrangement will be avoided tilt problems.

What is the principle behind your thought ? If I had to guess, I'd say the current (torque) distribution of one VFD would happen on the whole ring rather than only a section of it. Problem with Figure 2 being that the leveling out on the VFD level (as opposed to motor level of figure 1) is not possible because each VFD would only work in V/f mode and interconnecting them woudn´t help. Am I right? (I am an newbie in VFD and load sharing, i'm just making wild assumptions here ).

 

[iop995]

Yes, like fig. 1; in this scheme, load is equally distributed to each hoist and if shared load is below maxim hoist (induction motor) load, speed will be very close each others. Maybe need a control scheme with acustic/light signaling to alert when a motor don't run.

 

[LionelHutz]

A pulse train from the PLC to each VFD will control the speed very accurately and allow the speed of each VFD to match so all motors see the same frequency. With an accurate frequency match, fig 1 vs fig 2 vs a VFD per winch should all perform equally well.

Using 4-20mA speed signal could introduce enough error between each VFD frequency that the distributed method might provide a more even lift.

 

[Skyba]

Maybe need a control scheme with acustic/light signaling to alert when a motor don't run.

Yes, I'll try to link the thermal overload to the emergency stop of all the VFD to shut down the system if any motor fails.

Using 4-20mA speed signal could introduce enough error between each VFD frequency

Here again, i'm walking in darkness. The PLC->VFD current plays a role too ?? Aaaww so many parameters !! How can a beginner possibly have knowledge of such a thing... Can you tell me a bit more ?

If you happen to know these items: I am thinking of using a "Siemens LOGO!" PLC, with 1 to 3 "3HP SEW Movitrac MC LTE B" or "MC LTP B" variator, along with a set of 0.5 HP "SEW Motoreductor R17 DRS71S4".

 

[ozmosis]

Skyba
Some other considerations:
What is the motor cable distance from the planned VFD and motors? It is important to know that maximum cable length achievable with a VFD would be calculated as each motor length summed together. Too long for the drive rating and you will have problems.
Usually, with a VFD running a hoist application, the VFD will monitor the load on the motor to ensure the timing of the brake release is coordinated with the run command on the drive. This prevents over-current of the VFD trying to start into a motor that is already braked and also preventing the brake releasing before the VFD has generated the correct torque in the motor (I.e the load could drop). With multiple motors on hoists (that will have a brake), I struggle to see how this can be coordinated. How does the VFD know when each motor is ready? the first motor to release the brake will tell the drive to run but it may cause problems with the other motors.
I see some challenges with this setup.

 

[Skyba]

the first motor to release the brake will tell the drive to run but it may cause problems with the other motors.

Actually, I do not need any brakes, because the hoist has an auto locking mechanism =). And even without locking : The force that is transferred to the chain connected to the motor is solely used to lift or lower the tank. The weight of the tank is only and directly transferred to the hoist hook.

 

[Compositepro]

I suspect that your boss will look at this proposal and stay with the traditional manual method for good business reasons. Construction environments are very tough on equipment, particularly electronics in the rain. How often would the system be used? How long does it take to set-up compared to how long each use duration is? Who will maintain it and operate it in two years. Where will you store it between uses? Concern about OSHA and UL issues.