How to determine servo motor torque sizing

[LeeG]

I am building a scissor style jack mechanism that will operate using a 1/2-13 screw. I would like to operate the screw using a 12v servo motor, and am not sure how big of motor I need.

The mechanism is a 2 level, rhombus style mechanism, with the screw at the bottom. The lift will need to handle at most 200 pounds. Very light duty - it will only be used a few times a week for a single up and down operation.

The total travel on the screw will be about 12". At 13tpi, that is 156 revolutions to completely open or close, and I don't want that to take more than about a minute, so 150-200rpm sounds right, but what torque should I be looking for? I am sure there are formulas somewhere for calculating stuff like this, but I am unsure as to what specifically to search for, so any help would be useful.

Here is basically what I am building.

View media item 105164
If someone has another option to get my to the same destination, I'd be happy to hear it. I have about 12" of space below the platform when it is closed, and need to raise it about 24" higher.

Thanks,
Lee

 

[Dustball]

Not going to answer the torque question but I will say that you're going to want to use an acme lead screw and matching nut for this, 1/2-10.

https://www.mcmaster.com/acme-threaded-rods/acme-lead-screws-and-nuts/

 

[nadogail]

My recommendation is to use a Ball Screw and Ball Nut. They are more efficient than a threaded rod and nut.

 

[klxrcr]

If you are applying force horizontally to the bottom leg, the required force when the scissor is closed can be very high. Ignoring the weight of the scissor mechanism, with 200lbs on the platform and the legs at 10 degrees from horizontal you would need 1130lbs of force to lift the load. Most scissor lifts will have the the actuator positioned between 2 legs to lower the force needed to lift from the closed position.

 

[Firebrick43]

My recommendation is to use a Ball Screw and Ball Nut. They are more efficient than a threaded rod and nut.

No, you do not want a ball screw for a load lifting application. They have such little friction that the load will back drive them and cause the load to fall unless you add a brake which adds much to the complexity and therefore negatively safety.

Also in a 1/2 size you are only going to get 3 or 4 groves per inch on a ball screw which means you loose reduction and therefore need a larger motor compared to say a 1/2-10 acme. On top of that you have to keep the screw protected and clean as ball screws can not tolerated trash what so ever causing them to lock up. The only two reasons to pick a ball screw over an acme is you need a zero back lash solution or you need really fast speeds.

To the OP, I don’t think you really want a “servo” for this application. Servos are expensive due to their fast speeds, encoders, and complicated drives/control mechanisms. To get torque out of a servo you need to exchange speed for torque with a reduction unit of some sort such as a planetary gearbox or timing belt. And there is no good reason I see for the positional accuracy an encoder affords for a lift.

I think you really should be looking at linear actuators with limit switches for your application. If you still want to build the whole thing a dc gear motor with limits switch’s, driving an acme screw and a pwm control would be they to go(you just built a linear actuator) A large stepper motor would be a more positionally accurate if you need that but much cheaper and higher low speed torque than servos. (They just don’t have high speed torque or know if they have missed a step).

 

[Innovate1]

Sounds like about the size screw found in many auto jacks. You could use the acme thread from one of those which work very similarly. Pick up one cheap from a junk yard.

For finding the torque needed put a lever on the shaft - for example say 12 inches. Then load the platform. Use a spring scale to determine the force at 12 inches needed to lift the load. Torque is arm length times force. Add some safety factor for aging and extra torque for higher speed and you have your answer. I would go with about 1.5 or 2 times the torque you measure to give a good margin.

You could also use a torque wrench on the screw shaft to find the torque.

Limit switches would be a good idea unless the motor is protected at stall and the force is low enough to not damage the mechanism.

 

[Trapps]

https://www.servocity.com/motors-actuators/linear-actuators/super-duty-linear-actuators


This one specifically

Not sure if this is a fit. I have sourced several items from Servo City and they have great product and service.

There are also heavy duty planetary gear motors you might look at too. Choose a speed and torque combo to fit your needs. They also have any mounting components you might need...

https://www.servocity.com/motors-ac...ar-motors/5202-series-yellow-jacket-planetary

 

[Innovate1]

https://www.servocity.com/motors-actuators/linear-actuators/super-duty-linear-actuators


This one specifically

Not sure if this is a fit. I have sourced several items from Servo City and they have great product and service.

There are also heavy duty planetary gear motors you might look at too. Choose a speed and torque combo to fit your needs. They also have any mounting components you might need...

https://www.servocity.com/motors-ac...ar-motors/5202-series-yellow-jacket-planetary

Trapps, those gear motors look really interesting and good prices too. Thanks for that! The linear actuator would work but is total overkill. The OP stated his lifter was for 200 lbs max.

 

[Dustball]

Not sure what the actual application is for but how about modifying a lab jack to add a motor to it?

 

[ClappedOutBport]

Not going to answer the torque question but I will say that you're going to want to use an acme lead screw and matching nut for this, 1/2-10.

https://www.mcmaster.com/acme-threaded-rods/acme-lead-screws-and-nuts/

I came here to say exactly this lol. 1/2-10 acme will last much longer and take less torque to drive.

 

[LeeG]

Thanks for the input guys. The end goal is for a lifter to raise the bed I was writing about in this thread.

My initial attempt didn’t constrain the scissor mechanism in the y axis, so it bent rather than lifted. I can accomplish the same thing as a screw by fixing the actuator can only move in the x axis. I just need enough rigidity in the lift so it only can rise in the z axis.

Lee

 

[ddawg16]

Math.....now you know why you needed it.

All the stuff in that is math.....Moment of Inertia, etc.

For 200 lbs+? I'd start to look at hydraulic. Unless position is super critical

Those Acme screws are not exactly cheap.

Servo motors are also not cheap. Not to mention the drivers. If you go servo, look into gear reduction. A smaller servo motor with reduction drive could be a lot cheaper than a larger servo motor...and gives you greater control.

 

[laser3kw]

Math.....now you know why you needed it.

All the stuff in that is math.....Moment of Inertia, etc.

For 200 lbs+? I'd start to look at hydraulic. Unless position is super critical

Those Acme screws are not exactly cheap.

Servo motors are also not cheap. Not to mention the drivers. If you go servo, look into gear reduction. A smaller servo motor with reduction drive could be a lot cheaper than a larger servo motor...and gives you greater control.

Usually you approach the solution from the desired result backwards. You need to solve the power required to lift 200 lb 2ft vertical (against gravity) in under a minute. Then you need to put the mechanics in to elevate the load taking into consideration the loads center of gravity, moment of inertia. Then you put in the prime mover (power unit) to actuate the system. The OP has a rough idea what he needs, he just needs to work out the kinks.

 

[Innovate1]

Math.....now you know why you needed it.

All the stuff in that is math.....Moment of Inertia, etc.

For 200 lbs+? I'd start to look at hydraulic. Unless position is super critical

Those Acme screws are not exactly cheap.

Servo motors are also not cheap. Not to mention the drivers. If you go servo, look into gear reduction. A smaller servo motor with reduction drive could be a lot cheaper than a larger servo motor...and gives you greater control.

You are right about the math technically. But practically there are a number of efficiencies, coefficient of friction and such, that must be estimated. What are the losses in the lead screw? The loss in the reduction (if there is one)? So just measuring it is more likely to give an accurate result and is easy. Less chance of errors too.

As for servo motors - not required. Just a gear motor with limit stops. They can be found for much cheaper.

I still say a trip to the auto junk yard is in order. Possible items - seat positioning motors, jacks for lead screw, air lift cylinders from a hatchback to offset weight of load,...

Looks similar to a murphy bed. Might want to look at how the mechanism on those works.

 

[matt_i]

The unknown is the torque required to turn the screw thru the nut under a given axial load. There's a theoretical torque based on clampload in a fastener but it the generic "fudge factor" has to be refined for experimentation based on lubrication and the "fit" of the fasteners.

What I would do is setup your structure and use a regular torque wrench or a spring "fish scale" on an arm of measured length to get the actual foot pounds needed to drive the screw. As noted the max load will be found when trying to launch the mechanism from its fully collapsed position.

You will also figure out quickly if you need a brake. If you can let go of the arm then its all good. If the arm starts to self-rotate at any position (like without the weight of the arm itself acting via gravity) then I'd recommend one.

It could be as simple as a hinged device which mechanically limits the travel of the scissor or even an extra link which triangulates the mechanism into a ridgid structure that one applies at max travel/full height rather than a brake which would hold at any desired height.

 

[cgall]

Much cheaper to use an air cylinder if it has only 2 positions.

 

[LeeG]

You are right about the math technically. But practically there are a number of efficiencies, coefficient of friction and such, that must be estimated. What are the losses in the lead screw? The loss in the reduction (if there is one)? So just measuring it is more likely to give an accurate result and is easy. Less chance of errors too.

As for servo motors - not required. Just a gear motor with limit stops. They can be found for much cheaper.

Looks similar to a murphy bed. Might want to look at how the mechanism on those works.

Doing the math wasn't an issue - knowing which math to do (i.e. mechanical engineering) is.

There has been some good input so far. Some of my issue is that I don't know some of the terms. In my mind, small, low speed, high torque motor = servo. I don't specifically need a servo, just something that fits, so now I know to look for gear motors. I have used acme threads for other projects (even cut them on my lathe), but didn't think it was necessary for such a small weight. 2' of 1/2" acme all thread is less than $10, so it isn't a cost issue.

The bed already has a murphy bed style lift mechanism. The problem is that it isn't quite strong enough to lift our heavy mattress. I am looking to augment the existing mechanism with a powered lift. I've never done anything like that, and I figured it would be a good learning experience as well has being a cool end product. Most of my work is with static structures rather than dynamic, so this has been a good learning experience.

I appreciate all of the input.
Lee

 

[zhaddock]

https://www.amazon.com/Progressive-Automations-Linear-Actuator-Stroke/dp/B00OTX4RGC/ref=sr_1_1_sspa?_encoding=UTF8&c=ts&dchild=1&keywords=Linear+Motion+Actuators&qid=1594655280&s=industrial&sr=1-1-spons&ts_id=350654011&psc=1&spLa=ZW5jcnlwdGVkUXVhbGlmaWVyPUEzUUVYUk9XWElSU0UzJmVuY3J5cHRlZElkPUEwMzc5MDIyWFNIRDhQUFc5RUlOJmVuY3J5cHRlZEFkSWQ9QTAwMDMxMzdRR0czNEJUVURCOVcmd2lkZ2V0TmFtZT1zcF9hdGYmYWN0aW9uPWNsaWNrUmVkaXJlY3QmZG9Ob3RMb2dDbGljaz10cnVl

 

[ClappedOutBport]

The main issue with scissor jacks is that down low like that they have nearly zero mechanical advantage. So, if they have to lift the the entire load starting bottomed out, you're going to need a jack which is multiple times beefier than it needs to be. As far as doing the math. Take the sin of one of the link angles when it's bottomed out, and when it's at full height. Reference the two against each other and you'll see how bad it is.


Eg: With a 5 degree resting angle and 45 degree raised angle
Resting: 0.0871
Raised: 0.707


So it will take approximately 8 times more motor to lift it at the start, then when fully raised.

If you are applying force horizontally to the bottom leg, the required force when the scissor is closed can be very high. Ignoring the weight of the scissor mechanism, with 200lbs on the platform and the legs at 10 degrees from horizontal you would need 1130lbs of force to lift the load. Most scissor lifts will have the the actuator positioned between 2 legs to lower the force needed to lift from the closed position.

Edit: I missed this post saying exactly the same thing I am. klxrcr probably explained it better anyway.

 

[Dustball]

Firgelli is a linear actuator manufacturer we've used before at work.

https://www.firgelliauto.com/

I'd rethink the scissor lift design and go with two horizontally mounted linear actuators at the base, one on each side of the bed. There would be a bearing/roller at the actuator/linkage connection to take the vertical weight.

If you use this in combination with your existing lifts, there shouldn't be much effort required.

 

[kbs2244]

I would look into a curbside treadmill
they use what you are talking about for the incline adjustment