Rolling Loads - Machinery Skates

[kartracer087]

Hello,

So I am not a structural engineer but I know enough to be dangerous I suppose and I had to take some structural engineering courses in my undergrad.

Anyways, say that I have a load that weighs 35,000lbs and it is to be placed onto (4) machinery skates that have rollers. Those machinery skates will be placed on top of timbers that are wider than the skate themselves. The timbers rest on a continuous concrete floor (i.e. the bottom of the timbers is continuously supported) and are only used to help elevate the load up to be flush with another concrete pad at the end of the skate's "route".

Here's where it gets interesting and I think I understand:

Each machinery skate theoretically divides the total load by 4, so that means I have a load of 8,750lb on each skate. But each skate has rollers on it. If the skate has 2 rollers evenly positioned is it fair to say that in effect this creates (2) separate point loads? So each point load is roughly 4,375 lbs. And the issue here seems like this is sort of like a really short column loading issue with the wood, correct? Is it safe to say that the 2 point loads if even can be thought of as the original 8,750lb load distributed over a face area that is equivalent to a square with the outer edges bound by the point of contacts of the skates. I.E. if the skates are 8 inches wide and spaced 8" apart, then I have a stress area in the wood of 64 square inches which I can't exceed, so for the timber lets say crushing stress is 1,000 lb/in2. In this case my calculated local stress at each skate is 8,750lb / (64in2) = 137 psi which is acceptable given the compressive strength of the timbers.

Locally the stress would be much much higher at each skate since the contact area would be tiny (looking like a line basically). However, I think that becomes irrelevant since there is in effect a point load at the geometric center between the skates and the force distributed over that area proper since the entire bottom of the timber is being uniformly supported by the concrete floor. Which based on my research means my timber is acting like a tiny column in this case.

Is all of this essentially accurate and are my calculations in order with what would be expected? Is there any concern over the localized "line loads" imposed by the rollers? the stress at these points would theoretically be very high but I'm not sure that is cause for concern. Isn't this related to Saint Venant's theorem essentially? Even though when we look at the loading we are only taking about a few inches of separation between the hypothetical center of load and the actual pressure points the skates make on the face of the timber.

Thanks

 

[EdStainless]

Will this be possible to do?
Or will the local contact pressure from the rollers deform/indent the wood so much that it will prevent them from rolling?
Your loading will be on a line that is only thousandths of an inch wide until the wood starts to deform.
If we had a steel plate on the top of the timber for the rollers to run on, then the loading would be dictated by the stiffness of the timbers.
I have seen such rollers tear up (flaking and spalling) concrete floors due to localized loads.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[human909]

TIMBERS!? What sort of soft skates are you using!?

Last year I designed and performed a move of a 17tonnne (37,500pounds) piece of equipment. I used 4 skates. I absolutely did not use timbers which as EdStainless mentioned would be problematic. I design the rolling surface to be suitable for point loads of 50% of the total mass. Designing it for 25% of the total mass is quite unconservative, while 50% is conservative, I didn't feel this was a situation to muck around with. With 4 supports you can theoretically get 2 supports taking the vast majority of the load.

I also had to add joists and add temporary 12mm plate cover on a steel floor. This is what was required with the above assumptions. While it might have been somewhat conservative, the job was completed seamlessly in half the time allocated. Which was a massive win considering the cost of equipment mobilisation.


TL DR; Don't muck around with heavy moves. Design suitably and conservatively and you can get it done efficiently.

Is there any concern over the localized "line loads" imposed by the rollers? the stress at these points would theoretically be very high but I'm not sure that is cause for concern.

Absolutely. As EdStainless said, rollers can readily damage concrete flooring. Timber would naturally be far worse.

 

[SWComposites]

Yeah, just put steel plates on top of the timbers and be done with it.

 

[kartracer087]

Ok makes sense so it’s not necessarily the “column loading” compressive strength that is the issue it’s more of the deformation/deflection caused by the roller point loads moving on the timbers so adding a steel plate will redistribute the load on the joists so the rollers don’t indent the timbers while rolling. Makes sense actually.

This was purely a hypothetical I’m not moving anything just considering how one might move a very heavy object in a tight space where it may only be possible to crane lift the load through the door opening and then you would need to roll the equipment the rest of the way in the room into its final permanent location.

I know they also make nylon rollers too that are a bit softer but the issue is for them to be used on wood they would have to be pretty soft to spread the load out..

The other thing might be to use 12”x4” A500 rectangular channels of adequate wall thickness instead of the 12x4 rectangular timbers.

Thanks for answering my question!

 

[human909]

Yep. I'm now about to do my second job moving a >10T load in an area where cranes cannot access and the floor isn't strong enough.

First one we strengthened the floor and added plate everywhere. On the second one we couldn't strengthened the floor but we could do it in 2 perpendicular moves. So we we are laying 'rail tracks' of beams with up turned channels on them for the skates to run in.

 

[bones206]

I’ve seen some impossibly large equipment moved with air skates. It’s quite surreal to watch.

 

[bones206]

Here’s a good one:

https://youtu.be/4Th6YEQjBO4?feature=shared

 

[kartracer087]

So here is a conceptual drawing of my last comment, with the HSS steel channels being used with the skates to move into place, for reference. Stacked solid plates could also be used to get the 4" elevation rather than the HSS but they would be heavier. Nylon rollers would be needed so it can roll smoothly on top of the housekeeping pad without damaging the concrete.

There is a video here that has a somewhat similar process although the transformer is much much heavier and off loaded by a rail car or truck: Link

 

[bones206]

I don’t think hollow tube sections would be a suitable track member. The wheel loads could squash the loaded face of the HSS and buckle the side faces.

 

[dvd]

I would start with (2) stacked 2x12's of douglas fir with a C6x8.2 toes up (if it clears the lowest part of the caster bracket and accommodates width, depends on caster dims). Could also use a piece of bar or plate. Probably should fasten it all together. If you are just thinking about this you could mock up a small section and load it in a hydraulic press to see how much deformation you get. You would be just shy of a 4" stack up with this - caveat emptor.

 

[kartracer087]

DVD I think that is a neat approach because the C channels would help make sure the skates stay on the rails that is one big downside to what I proposed is that you would really have to be careful when moving it to make sure you stay aligned and on the surface of the HSS.

On another note, I did do a quick calculation figuring HSS with a center point load and supported at 60" intervals with load at half the support distance. Actually in my case the support interval is 0 since it lays flat on the floor. The stress calculated for supports 60" apart and one point load of 7,500 lbs was around 7,000psi and the maximum compressive strength of the channel is much greater than that. My wall thickness I used was 0.375" so its pretty sturdy. But you are correct bones206 because I looked at some research papers and yet again the rollers cause too much localized stress and so you can definitely get buckling of the walls. So I think you need plates to even out the load no matter what you do, i.e. you can't really just roll it over a section that has in effect 2 walls. A solid steel plate is the best bet by far because it doesn't have the deflection issues. Simple analysis of something like this may not be sufficient you may need a program that can do FEA if you really wanted to get an accurate gauge of the stress. Or as noted, a C channel laid flat with the channel side facing up acts like a bearing plate in effect.

Could you still do this with HSS? Sure but you need a much lighter load that isn't going to push down so hard. This piece of equipment is ridiculously heavy and gives (4) concentrated point loads so its very tricky. One thing this has shown me is rigging costs to move something like this get very expensive. Say your rigging company doesn't have the steel to do it well you could be paying 5-10k just for the temporary steel to do the set up. Its wonderful if you can fit a crane as that makes it a lot easier and you don't need to be concerned about roller stresses on surfaces but there are some cases where you just can't get a crane in there, even a small spider crane. And those have load limitations as well. So its kind of a lessons learned by a non structural engineer who deals with extremely heavy equipment (well specifically, transformers, generators, and switchgear apparatus). The only people who kind of know these pains are the mechanical guys moving boilers, tanks, and chillers. Its sort of like "hey I've got a 5 car stack of G wagon SUV's sitting on top of each other, is this ok to support? Lol.

The more you think about it there really are only a few safe ways to do something like this given the weight of the load and the types of construction involved. Sure in a power station for example they might have very heavy structural walls and movable beam cranes to move things but in a pretty typical building you aren't going to have anything remotely like that.

And yeah this is a hypothetical case. I was only doing it to sort of think about what's actually involved and some areas where you might encounter red flags. A professional rigger should know all of these variables but either way it does help me understand some of what may be involved to talk with some knowledge of what is actually needed for a job like this.

 

[human909]

Yep, as I said above I'll be using the upturned C-channel method soon, likely within a month.

Watch out for jacknifing. I found there was a narrow window between too tight and likely to give problems and too lose and able to jacknife and get caught. The skates should be tended so this doesn't occur, but it is a risk you don't want to take.

I found the sweet spot by "playing trains" in the workshop. For a 200x100 skate. 130mm (110 between radius) was too tight. 175mm (165mm between radius) was too loose and allowed jacknifing. 158 (145mm between radius) was perfect.

You could probably work this out by hand based on the allowable angles of the skate. But it was more fun getting my hands dirty and playing trains.

 

[dvd]

Stress is not as relevant as deflection. You don't want to have the roller in a deep valley and constantly be pushing it up out of the dip. I have seen too many massive loads on timber rafts to not have confidence in a wooden base for this activity. You need the steel track to avoid excessive deformation of the wood under the roller line loading.

 

[BridgeSmith]

Each machinery skate theoretically divides the total load by 4, so that means I have a load of 8,750lb on each skate.

that assumes the loading is centered transversely and longitudinally on the skates, and the force to move the machinery is applied low enough to not create a significant moment.

for the timber lets say crushing stress is 1,000 lb/in2.

It seems you've been talked into using a steel track, but if you're still going to check the pressure on the wood below, make sure you're calculating the capacity of the timber using the correct allowable stress. If I understand correctly the timbers are laying horizontally on the floor. If that's the case, the stress limit for the timbers would be for compression perpendicular to the grain (crushing the side of the timber), which is much lower than for parallel to the grain (crushing strength for a post/column loaded on its ends.