Belt tension on conveyor gallery structure

[RattlinBog]

Looking for guidance on a few things below. See attached pdf. I'm analyzing an existing conveyor gallery and trying to apply belt tension correctly. My head and tail pulleys are supported outside the gallery (at a transfer tower and separate building, respectively) so I believe the gallery will not be subject to total tension (T1) from belt loaded side. That should be supported at the other building/tower. However, the gravity take-up bend pulleys are supported by the gallery truss steel, so the gallery will see slack side tension (T2) from the bend pulley reactions. I'm not a conveyor expert by any means, so perhaps I'm off somewhere. I reviewed the info here for belt tension reference: [URL unfurl="true"]https://rulmecacorp.com/how-to-calculate-conveyor-belt-tensions/[/url]

1) I currently have my truss supports as pin-pin. I don't think that's quite right. I'm thinking of keeping the top-right pinned and change bottom-left to roller. Bottom-left chord is bolted to the web of a W8 column, so I imagine it's not stiff enough to be a true pin for restraint in x-direction. Does the proposed change sound correct? Connection details are on pg 5 of pdf.

2) Does my FBD for the take-up bend pulleys look correct? Am I missing any other belt tension-related forces on the gallery?

3) I have dimensions for the counterweight, so I know it weighs 2250 lbs. Per original drawings, the design slack side tension, T2, is 5539 lbs. For equilibrium, my counterweight would have to be 2*(5539 lbs) = 11k...which I know it isn't. So in reality, are the take-up bend pulleys only seeing loads based on the real counterweight weight?

Thanks in advance

 

[human909]

You seem to be on the right track.

1. For any long trusses I'd usually use pin and roller. Often pin, pin doesn't make a big difference, but sometimes it can. Especially when there are two connections as they will form a moment connection if rollers aren't used. I find this with lateral (eg wind) loads. Also for long trusses I will often explicitly allow movement in the connection at one end.

2. Your first FBD looks right. Your second one has the belt tension at horizontal as opposed to in line with the belt.
Your second FBD also has horizontal forces that don't net out to zero, so recheck your results.

3. Your logic and conclusion seem correct.

 

[RattlinBog]

Thank you! I made some adjustments and updated my risa-3d model (see attached). This looks more reasonable. It's pin-roller now; I changed the panel point loads to be vertical; fixed bend pulley loads to be in-line with belt; and cut bend pulley loads in half since I'm looking at only one truss. (Axial force: green = compression, blue = tension)

 

[human909]

You likely need to reduce the pully loads by 75% if what you have stated is correct and I'm understanding it correctly.

Counter weight = 2.25kip
/2 for each side of the belt
/2 portion per truss (2 trusses total)
=0.56kip

Somewhere along the line somebody either got their conveyor tensions weight wrong or the reported tension is wrong. On or the other.

The other thing that is missing is roller friction in the direction of the conveyor. While it is likely low, it is still there and it is distributed load across all the rollers acting on the truss in the direction that the conveyor is moving product.

 

[dvd]

Are you, as the structural engineer, calculating the belt tension? Make the conveyor designer give you that information.

Is your counterweight weight calculation based on concrete, or steel plates, or punchings, or...? The conveyor won't run when there is insufficient slackside tension, so probably a good idea to get the counterweight and tensions agreed upon before someone makes changes to the loadings.

 

[RattlinBog]

human909, you're right that I'm currently applying a larger load than is likely there. However, I'm trying to be a bit conservative as I'm checking the gallery steel due to some corrosion issues. I'm not 100% confident on the counterweight.

dvd, no, I'm only back-calculating the belt tensions that are already given on a conveyor design data sheet from 1976. Te, T1, & T2 are given. I just took the motor nameplate HP divided by belt speed to get Te to double-check the given Te. The numbers matched (see attached). I felt like I could trust the design belt tension values after that. To be clear, I'm not designing the conveyor or any mechanical systems here. Just structural analysis of an existing gallery.

Counterweight weight was based on concrete dimensions of 1'-1 1/2" x 2'-8" x 5'-0" x 150 pcf = 2250 lbs. However, there may be more steel plates, etc. that I'm unaware of. Since I'm just focused on checking the capacity of the truss steel, I'd rather be a bit conservative on my loads, so I decided to use the T2 value instead of the counterweight concrete weight.

 

[RattlinBog]

Forgot to attach my hand calcs for effective tension

 

[dvd]

RattlinBog - kudos on nicely canceled units in your calcs. You officially appear to have your head on straight.

 

[canwesteng]

Where does the w8 connect to on the roller side? Pretty sloppy detailing by the original engineer, is the conveyor going to bridge load from one structure to the other?

 

[RattlinBog]

canwesteng, see attached elevation that I drew up. The original drawings aren't too bad; you just have to look at some sections and details to get the full picture. I also have the set of erection and shop drawings for reference.

The W8 is part of a bent. The main conveyor gallery is spanning from the transfer house to a bent, and then there are a couple smaller-span trusses from bent to bent. The conveyor continues on (to the left in my elevation) to another building where it eventually ends at the tail. I don't have the transfer house framing fully drawn in my sketch, but the gallery connects to a W24, which is braced laterally at each chord connection. I believe that to be the pin end. The bottom chord connection to the W8 bent is on an L6x4x3/4 seat angle with standard holes. I believe that to be the roller end (even though holes aren't slotted) because it doesn't have enough stiffness. Top chord connections are slotted holes.

 

[RattlinBog]

canwesteng, attached risa model may help with visual too. I'm showing DL + LL reactions. Let me know if you think I'm off with my end conditions.

 

[canwesteng]

Ah, that makes more sense. The bent is considered pinned at both ends, so it should be considered a roller if you stop the model there.

 

[PersonalProfile]

I propose a competent person reviews in detail, though as start some considerations follow:
1. Consider a lateral allowance for drift/misalignment etc,
1a. Consider an allowance for the motor’s: starting torque, and/or break down torque etc, refer typical squirrel cage torque speed curve following:

https://avsld.com.sg/typical-torque-speed-curve-of-a-cage-induction-motor/

2. Consider including pulleys’ masses, and
4. Consider including mass of: conveyor belt, conveyor idlers, and conveyed material etc.