Ratio issues

[itsmoked]

I've built a stand comprised of four screw jacks ganged together. (Teach me not to wander from the electrical reservation!)

[unloaded]

I did meticulous empirical testing to discover how much force needed on a specific moment arm was required for one of these jacks to lift X pounds.

I rounded everything up and used the next size up of everything.

It is driven by a gear drive motor @ 60Hz that is rated at 10.15Nm @ 158RPM (250W)

[motor plate]

It goes up and down 'just great' taking about 2 minutes... Unless the design load is on it! The load is 3,000lbs.

My test load is 1.60 tons(3,200lbs) accurately measured on commercial scales.

[loaded]

The motor starts very, (very), hard then continues raising the load up the 18in distance while drawing 550W. Near the top ~15 inches it really starts bogging. Hitting 900W and and pulling 8+amps into the 120V voltage-doubling VFD just before stalling entirely.

First thought was that the four jacks are binding up tight and just bringing everything to screeching halt. A plan was hatched to test this theory. A sprocket had a nut welded to it and a bending-bar torque wrench was used to continue lifting after a stalling point with the motor was reached. The nut/sprocket was mounted onto the motor's shaft.

You can just see the nut/sprocket on the end of the motor shaft - the second sprocket.
[oblique]
[Larger image]

http://www.box.net/shared/static/okpj9legnb.jpg

No hard binding was detected. It just takes 19.5~20.0ft-lbs to continue raising the load. It continues up smoothly with no hitches at the 20ft-lbs. By hand, slow or fast, the 20 is constant.

This looks to me like (20 x 158)/5252 = 0.6hp ~ 450W which is pretty much being shown by what I'm seeing.

Presently the gear train includes the motor and its gear box. This is followed with a 19T sprocket driving a 60T sprocket. Using two other 60T sprockets the power is transmitted to the opposite side.

Their is no hurry as the system takes about 2 minutes to raise unloaded, five minutes presents no operational issues. My thought is to increase the ratio further.

[above gear train]

I'm looking for suggestions. Something that will fit the present geometry without major surgery. Something like 3:1 maybe. I'm seeing no suitable way without a bunch of changes.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[MikeHalloran]

It looks like you're applying a fairly serious chain pull to what I think is the jacks' input shaft, and it doesn't have the bearings for it, even neglecting where the sprocket overhang multiplies the moment.

I'd be inclined to drive the intermediate shaft, and support it in its own set of bearings, e.g. flange bearings or hanger bearings. The intermediate shaft could probably be a little fatter, to better resist bending from the sprocket loads. That's the second best part about Lovejoy couplings; within a series, jaws with different bores will interchange.




Mike Halloran
Pembroke Pines, FL, USA

 

[williedawg]

If the legs are not parallel to one another (they should be aligned with less error than the clearance within the tubes that make it up), in sync and the whole system not level, there could be some serious binding as the table is raised, especially with the "rated" load applied

If one or more of the legs extend at some angle to any other, it only gets worse as the height is increased.

And, I agree with MikeH, those shafts/fixed tubes should be bronze-bushed. An outboard flange to support the Lovejoy half that's common with the sprockets would also help. This could also be bushed. Obviously concentricity of the shaft/sprockets/Lovejoy is a must.

 

[MikeHalloran]

Without changing much else, you could maybe go a couple teeth smaller on the gearbox sprocket and a couple teeth bigger on the first driven sprocket. That might help some.

I agree with Willie; better alignment will help more.
The behavior suggests a dynamic alignment problem.

Right now, the front legs are laterally braced in only one direction.

E.g. you've got X bracing between the 'rear' legs in a vertical plane... but you might add X bracing, or something structurally equivalent, in a horizontal plane, to prevent the 'front' legs from going bowlegged or knock-kneed.



Mike Halloran
Pembroke Pines, FL, USA

 

[BobM3]

How torsionally compliant are the couplings? Possibly the two "direct drive" ends are moving more than the ends connected to the couplings (causing the binding).

 

[itsmoked]

There are four Lovejoys. The jacks are bronze bushing'd.

I have also been concerned by the alignment. But since the telescoping clearance seems to be pretty loose it would seem if one tube was going too splay too much the others would 'have some to give back'. Indeed, my hand torque test shows that at the bottom it's about 17ft-lbs to lift and at the top only 3 more ft-lbs. So there is probably some increased friction happening though not any nasty binding.

I'm surprised that no one makes two, mated, different diameter, bearing'd, idler sprockets on a short shaft with a single end mount. One could then interject one into a train to easily add a ratio change.

Thanks guys.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[desertfox]

hi itsmoked

Apart from the alignment issue's raised by others which are valid points I wondered about the distribution of the load ie how central is the centre of mass of the load with respect to the four lifting legs? its seems to me that some increased friction or binding can occur during the lift with an offset load.
The other thing that caught my eye was I couldn't see how you tension the chains and finally I am no electrician but how long will the motor rated at 250W last if your required power is actually double.

desertfox

 

[MikeHalloran]

Slack in the cross- chain and the couplings will cause some of the legs to lag, but I don't see that as an issue while the parts are still new.

I was thinking of ball bearings, not bronze, everywhere possible. ... including on both intermediate shafts; Lovejoy couplings are not really meant for supporting shafts.





Mike Halloran
Pembroke Pines, FL, USA

 

[itsmoked]

The continuing saga..

desertfox; How long will a 250W motor draw 700W? Not long...
And that load was centered with a ruler.
The chain is tensioned by using links/halflinks/offset links.

Mike new bearings everywhere?!? Gaww!! These jacks are each rated at 3500lbs. They had long handles on them to be cranked by humans. Every day. I would hope they could take my slower machine cranking once a week or so with the side thrust partially shared on both ends of the shaft by the neighboring jack.

During further testing the first driven sprocket was found turning while the secondary sprocket feeding the left side wasn't! It's hard to find sprockets with key-ways for 1/2" shafts. So I drilled the two gears and tied them together with roll pins.

Of course you know what happened next.. Both the tied together sprockets now spun on the unkeyed shaft.




The Lovejoys luckily are keyed. By tying the two sprockets to the Lovejoy with a TIG welder the sprockets are now essentially keyed. They do not slip now.

But back to the over current...
To drop the motor load I came up with an interposing gear reduction. A 20T/72T pair of sprockets welded coaxially.
Held on a support plate, on a precision shoulder bolt riding on two needle bearings. The motor gearbox runs the gear reduction large gear and the small gear feeds the original driven gear. The motor now runs at 205W meaning it isn't even fully loaded. The down side, and it isn't much of one, is that a cycle takes 9 minutes.

Everything seems OK except for one troubling detail.
Look closely at the above picture. Look at the chain. Look at what some of the links are doing. If you let the chain run thru your fingers it feels lumpy in various places. This is brand new chain. What causes that? It's only on that chain. The one with the highest load.. I have a bad feeling..






Keith Cress
kcress -

http://www.flaminsystems.com

 

[MikeHalloran]

It's normal for new roller chain to be a little stiff.

That said, this would be a good time to drag out the engineering section of the catalog and review your sizing calculations.



Mike Halloran
Pembroke Pines, FL, USA

 

[diskullman]

Use taperlock sprockets. They are available with keyways for 1/2" shafts

Russell Giuliano

 

[MikeHalloran]

Go to a motorcycle shop and get some aerosol foaming chain lube. It will work a lot better than grease.



Mike Halloran
Pembroke Pines, FL, USA

 

[itsmoked]

Thanks gents.

What happens if you overload chain? Does it deform or just break. I was thinking the rollers could be bent and that would cause what I'm seeing. Especially since it only appeared after that particular piece of chain had its load increased by the gear-down.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[MikeHalloran]

The roller chain size that I think you're using is of simplified construction. I'd expect the pins to be the weak link. They will bend a little before they snap.

I keep forgetting you're a sparky. That's a compliment, but it also means you probably haven't read the engineering section of the roller chain catalog, what with chains being so simple and all..

So, okay, backyard engineering analysis: Since you've at least doubled the tension on that strand, and it's getting wonky, it would make sense to substitute a larger chain size, or double or triple up on that pass, i.e. duplex or triplex sprockets and chain, or just multiple sprockets pinned together. Don't try to get fancy and stagger the teeth; just line them up nice.



Mike Halloran
Pembroke Pines, FL, USA

 

[itsmoked]

Thanks Mike.

I decided to re-do the math for a sanity check on the tension expected to be applied to the chain.

If we recall the torque wrench test. 20ft-lbs were required at the original motor gear box drive gear. It's 19T and has a radius of about 0.775".

0.775" x (1ft/12")= 0.065ft diameter.

To transmit 20ft-lbs on a moment arm of 0.065ft we would need a tension on the loaded side of:
20ft-lbs/0.065ft = 309lbs.....

Is this too much for my chain rated at a working load of 114lbs? [flush]

Man! That calls for a chain of at least ANSI 40!!
What a disaster..


Keith Cress
kcress -

http://www.flaminsystems.com

 

[MikeHalloran]

Yeah, that's what I meant by "fairly serious chain pull ".

Industrial drive design tables are conservative; you might get away with a slightly smaller chain size. I wouldn't try; the difference is less than a warranty claim.

Mechanical education can get expensive, but so far you got off easy; only your wallet has been injured.





Mike Halloran
Pembroke Pines, FL, USA

 

[patdaly]

Keith, any mom-n-pop machine shop will have Dumore keyway broaches. That would have solved your keyway dilemma. If not, they are reasonably cheap to purchase and only require a small press to use.

It is possible you will still lose it with the TIG weld...