Factoring Elevator Loads

[mikek396]

When elevator contractors provide machine room loads to the structural engineer, these loads have been calculated by doubling the live loads (car, counterweight, etc) to account for impact. Should I be further factoring these loads or is this overly conservative?
One engineer I spoke with mentioned they don't further factor the loads because they are using ASD philosophy. IMO this still seems overly conservative via the omega factors used for ASD.

For reference, this broader question is coming up because I have come across a situation where a previous contractor used a channel in minor axis bending. It looked odd to me so I ran some quick numbers. The only way this passes LRFD or ASD is if I use the static load values instead of doubling for impact. When using the impact values, it fails ASD by 40%. This elevator has been in normal service like this for 10+ years, with no obvious signs of deflection on the channel in question.

Any insight here is greatly appreciated.

 

[canwesteng]

The loads should be factored on top of the impact. I wouldn't carry those loads past the supporting members and into in the main structure necessarily, since they are instantaneous loads with a low probability of occurring at the same time as other design loads. For what it's worth, if something fails in ASD by 40%, it isn't likely to actually fail under that loading. You'd need to be 50%+ over to actually hit the nominal capacity of the member.

 

[human909]

For passenger comfort elevators are designed to accelerated at fractions of g. A quick google suggests 1m/s^2 which seems reasonable. Passengers don't like suddenly weigh twice their body weight as the elevator accelerates, a 10% increase is much more tolerable.

So the dynamic loads in NORMAL operation are likely only ~10% above static loads. And static limits are rarely reached and most modern elevators will refuse to operate if they are exceeded.

So it is entirely reasonable for the members to never see loads greater than 110% of the maximum rating of the elevator. (Safe design naturally should be a fair bit higher, I don't have the answer for that.)

I'm not sure why you'd double for impact. Is this a rule of thumb? Or just the mathematical result for a fall of a distance of ZERO? (which is 2x, the integration comes out quite nicely) But an elevator isn't impacting, they have strict controls on both acceleration and jerk. And if there is failures in controls the any "impact" could readily exceed 2x static load.

 

[canwesteng]

I believe the factor of 2 is coming from the supplier - probably a feel good number, but there are lots of things to account for, including the elevator getting stuck, any binding of the cable... The actual steel in the mechanical room is a pretty poor place to try and shave down the loads and steel tonnage, given the possible consequence. But again, I've never carried those loads down the structure (only worked with maintenance elevators on industrial plants fwiw)

 

[mikek396]

The 2x factor comes from the elevator code (ASME A17.1) which says that the supports must be designed for double the tension in the ropes to account for acceleration/deceleration (2.9.2.1).

Its true that during normal operation it is unlikely to experience the 2x, but this load is possible during emergency braking.

@canwesteng I agree the elevator supports are a bad place to try and shave weight - this only came up because I came across an existing situation that didn't sit right with me, and upon running the calcs I cant figure out how any engineer would OK this. If you wouldnt carry the given loads past the supporting members, what loads would you use if designing for a new elevator motor room?

 

[canwesteng]

Well, the members in the elevator room are the supporting members, so they are designed for those forces. If you find an existing elevator room with members that don't work for the rope loads, it's not unlikely that the original design had some errors in it.