# How much load is going through the bolts

1. Dec 20, 2013

### Stickyman

Hey guys.

Im building a special type of squat rack for my gym, and need some input on the bolts to use.

I have used high tensile M16 bolts on all joints, but am not sure if it is enough, hence me joining this forum and asking you guys.

Attached is a picture, and Ill explain the lines ive used it a bit more detail.

The white circles are M16 bolts.

The red arrow is a pivot point.

The yellow line is 880mm long, the pivot point to load.

The blue line is 1000mm long.
A Barbell sits inside the green slings. You squat up to 500kg, the bar could fall a maximum of 1000mm.

So, the question is;
If a weight of 500kg falls inside the slings, which is 880mm away from teh pivot point, how many kilograms of force will be on the 2 bolts at the pivot point? Is 2x M16 bolts big enough?

2. Dec 20, 2013

### Stickyman

3. Dec 21, 2013

### Baluncore

The answer to your question requires knowing how long it takes the falling mass to stop. That is determined by the flexibility of the structure.

If the structure has no flexibility and the falling load stops instantaneously, then the force on the bolts will be infinite.

4. Dec 21, 2013

### Stickyman

The link in my 2nd post has the video of my dropping the bar with 400lb from about 1m height.

5. Dec 21, 2013

### Baluncore

Then you can measure how long the mass takes to fall and how long it takes to stop.
I cannot access facebook and I avoid watching videos.

6. Dec 21, 2013

### AlephZero

I think you need some professional advice for designing this, since obviously a falling 500kg weight could cause a serious injury to the user.

As Baluncore implied, the force on the bolts will depend very much on how flexible the green (fabric?) slings are. Beyond that comment, IMO you need to do some proper calculations based on the full details of the structure, not just a photo and a couple of dimensions. Personally, I wouldn't like to guess if the bolts or something else would fail, and it would be a guess from the (lack of) information we have.

You could easily be talking about seriously high loads here. The multiplication factor from stopping the falling mass could easily be 10 or 20 times. Them you have a lever arm with a ratio that look like about 6:1

So your 500kg weight could be putting a load of 50 tonnes or more into the top of your frame. But that number is just a (plausible) guess - don't use it to design anything!

Last edited: Dec 21, 2013
7. Dec 21, 2013

### Stickyman

Sorry about the lack of information, I have no idea on engineering, which is why I'm posting here.

The green straps have 0 stretch in them.

I dropped 180kg from 1m into the straps, and guessed around 10cm of flex from the structure.

Okay, sorry man I've been at Ikea..

Basically, without knowing the angle of the member indicated by the yellow line, I've just had to do the calculations under the assumption that it acts perpendicular to the uprights.

The impact force of a 500kg object falling 1m = 49000 N

To find the moment force about the bolt you multiply that by 0.88m, which = 43120 Nm

Which is about 4397 Kg/m. So across 240mm that's giving me like.... 18,320 Kg of force distributed between the two bolts.

This is by no means the full story though, to give you an accurate measurement I would need to determine the compressive force on the front bolt and the tensile force on the back bolt, as well as whether or not they're in single or double shear. Which would still only give you a static result, from there you could then start looking at what's happening to the bolts and welds and **** dynamically.

Patrick, that's not plastic deformation, that's the "recoil" of the falling object. And 0.1m movement in the bar after impact seemed pretty reasonable based on Scott's test video. Plus, the lower that value is (distance traveled after impact) the higher the impact force becomes. So if you take it down to say.. 5-10mm the impact force value becomes something like 490^12kN

^^Completely agreed with Patrick on that. The rating of 8.8 grade M16 bolts isn't worth jack **** if they're subjected to multi-planar force (heavy **** stressing it in several directions at once), which a half-ton impact will certainly do. So if someone fails while going heavy, it's worth replacing them.

8. Dec 21, 2013

### nvn

Stickyman: In your video, I saw your system deflecting about 60 mm, not 100 mm like you estimated in post 7. I currently estimated 60 mm, unless you have more accurate data. Therefore, using a 60 mm deflection, I currently obtained a system stiffness of k = 704 250 N/m.

Are you sure you want to design the pivot point bolts for a 500 kg mass, dropped from a height of 1000 mm? When I checked this, the dynamic amplification factor was 18.0, and the M16 property class 8.8 bolt, located at your pivot point (red arrow) shown in post 1, was stressed to 380 % of the allowable stress, which exceeds 100 %, and therefore would be grossly overstressed. (And this assumes one end of your bolt has threads in the shear plane.)

The above is quite different from your test video, which used only a 182 kg mass dropped from a height of 650 mm. Therefore, what is the maximum mass and drop height you want the bolt to withstand?

9. Dec 21, 2013

### Stickyman

Awesome, that's such a great reply, thank you.

In real life, the most weight that will be attempted will be 380-425kg.
The fall into the straps in normally much shorter than in the vid I posted, I'm just trying to make it as safe as possible.

I could move up bolt sizes if needed.

10. Dec 21, 2013

### nvn

Stickyman: So would the maximum drop height be about 800 mm? Do you want to use ISO property class 8.8 bolts? Or property class 10.9 bolts?

Can you post the dimensions (mm) of your lug and clevis plates, at the pivot point bolt (red arrow in post 1)? From a front view, it will look something like my attached file. Dimension t1 is the thickness of your lug, t2 is the thickness of your clevis plates (one clevis plate on each side of the lug), and gp is the gap between the lug and each clevis plate (when the lug is centered between the clevis plates).

#### Attached Files:

• ###### lug-clevis-03.png
File size:
1,010 bytes
Views:
228
Last edited: Dec 22, 2013
11. Dec 21, 2013

### Stickyman

I can use what ever bolts I need to, but currently 8.8.

The plates are all 10mm thick, and there is 8mm clearance gaps either side..... A little to loose I know :(

12. Dec 21, 2013

### nvn

Stickyman: Your current gaps are terrible. I would not want those gaps to be more than about 0.5 mm, each. Is that something you could redesign?

Last edited: Dec 21, 2013
13. Dec 21, 2013

### Stickyman

I can close them up easy enough.

Will it make the structure that much weaker as is?

This whole thing has been made with a 5" grinder, drill press and a mig.

I don't have top gear.

Last edited: Dec 21, 2013
14. Dec 22, 2013

### nvn

Stickyman: For a 425 kg mass, dropped from a height of 800 mm, for the pivot point bolt, it currently appears you would need t1 = 24 mm, t2 = 12 mm, gp = 0.5 mm, and an M30 x 3.5, property class 8.8, bolt.

15. Dec 22, 2013

### Stickyman

Awesome.

Thank you very much.

If it's not a huge trouble, can you give me a brief run down on the calcs?

16. Dec 22, 2013

### nvn

Stickyman: That would be too lengthy and too detailed.

I want to point out though, the mass and drop height (and force) corresponding to post 14 is far more force than what you tested in your video. Even though the force in your video is much less, you can see, in the video, the structure looks like it is already being pushed to limits. I am doubtful your structure and weldments could withstand the far greater force corresponding to post 14. See what I mean?

17. Dec 22, 2013

### Stickyman

I do mate, I do indeed.

That's why I posted.

I'll redo the bolts as per your recommendation, and thicken t1 and t2, whilst moving the clearance to .5mm.

I'm also going to shorten lever arm ( yellow line) from 880mm to 600mm.

Thanks for your help mate, much appreciated.

18. Dec 22, 2013

### nvn

Stickyman: First let me mention, I think energy losses in this particular system are no more than 2 or 3 percent, and are therefore negligible.

Secondly, if you could measure the deflection for us, as follows, it would be appreciated. Place the empty steel barbell rod (with no disk-shaped masses installed yet) in the green slings, and measure the exact distance, in mm, from the floor to the bottom of each green sling.

Next, install any amount of disk-shaped masses onto the barbell rod, preferably a high mass, or fairly high mass (such as your 180 kg, or 350 kg). (And let us know exactly how much mass, in kg, you installed.) Now measure the exact distance from the floor to the bottom of each green sling.

If you could do this, and post this data, it would enable us to compute the system stiffness k1, and would be really interesting, and important.

Third, I think I now notice that your structure has a rear brace only on one side, where a hydraulic actuator is installed. This is quite different from what I thought yesterday, and it would cause the structure to be loaded very unevenly, causing a vast majority of the load to go to the side containing the rear brace. This makes it much more difficult (and inaccurate) to compute. If you are going to keep the structure designed asymmetrically like this, then we would need to rethink the answers, and it would become more complicated.

Last edited: Dec 22, 2013
19. Dec 22, 2013

### Stickyman

Hi nvn.
Yes, the pivot arm is only secured on the right side the bottle jack makes the top move up and down for height adjustments.

These are called monolifts, and are used in powerlifting competitions.

All the monolifts I've ever seen are built this way, which is why I half copied the design.
I do agree that the structure will be loaded very unevenly though.

I will be at the gym in 20 min, so can let you know the amount of deflection on the 880mm arm then.
I probably will reduce the lever arm to around 600mm though to help stiffen everything up. :)

Once again, I appreciate your help.

20. Dec 22, 2013

### Stickyman

With a 20kg barbell in the sling, it's 772mm off the floor.

With the barbell loaded to 410kg, it's 685mm

21. Dec 23, 2013

### munster

Hi Stickyman,

I searched "monolift" to learn more about how they are used and found that indeed they have that actuator on only one side. It's strange from a design loading point of view. But I now understand why your video showed such an asymmetrical outcome - the whole mechanism rotated to the left and the weights shifted to the left (relative to the man standing at the beginning of the video) because the left side flexed more than the right side.

I agree with nvn that it complicates all the load calculations - those are for symmetrical loading. But you could also assume that ALL the load goes through one side i.e. twice the load. It will oversize the bolt, but what's the cost of a bigger/stronger bolt compared with the collapse of the structure and serious injury?

Having said that, a key factor in the calculation is the falling distance. I understand that they are safety slings and they have to be long enough for someone to do squats. I would strongly urge you to use the shortest slings possible - that would minimize the load on the bolts and hence the size of the bolts.

But I also realize from the videos that this is an add on feature to the basic monolift design - it was never really designed to take falling weights. How about adding a separate arm to support the slings? This could be rigidly welded, instead of working through a pivot and it opens up more ways to design the support structure that could be a far stronger and more rigid arrangement.

22. Dec 23, 2013

### Stickyman

Mate, that is a perfect idea!

I will build it tomrrow.

I was thinking something like this, out of 100x100.

Suggestions?

23. Dec 23, 2013

### nvn

Stickyman: There is an old adage, "Measure twice, cut once." Are you sure you measured those distances in post 20 accurately? Did you check them twice? They are the static (non-moving) floor-to-sling distance with the 20 kg empty barbell rod, and the floor-to-sling distance after adding 410 kg of mass to the barbell rod, right? And these measurements are when the barbell rod and masses are gently placed in the green slings, not dropped, and are not moving (static).

If your measurements in post 20 are very accurate, then that is vastly different from what I estimated in post 8, and very different from your estimate in the fourth line of post 7.

If post 20 is correct, then it means the deflection of the system in post 1, in your test video in post 2, assuming a drop height of 650 mm, was 264 mm. I am finding that hard to accept, yet.

What is the center-to-center height of the concrete blocks in the wall behind the monolift? Is it 200 mm? If you watch the video carefully, I thought it looks like the distance from when the green sling is fully extended but unloaded, to when the sling is fully loaded, is slightly more than one half of a concrete block, which would be roughly 120 mm, right? But keep in mind, the distance on the concrete blocks is magnified, due to perspective, because the wall is further away. Therefore, it implies a deflection of perhaps 100 mm (?). But according to post 20, it would be a deflection of 264 mm. This is quite a discrepancy, so far.

On another subject, would you be able to post dimensioned sketches of your structure, in mm? Because your structure is asymmetric, we need to be able to know all the dimensions. We also need all the member cross-sectional sizes, such as 100 x 100 x __ mm, where the blank is tube wall thickness.

I am not quite convinced yet regarding fixed sling supports. They will still be highly loaded, and possibly will interfere with usage (?), and will make the structure more complicated (?). Sometimes, simpler is better, if you make it strong enough.

I generally would say, change the design based on stress levels, instead of changing the design before you know a few of the stress levels. You can always change the design after you know a few of the stress levels. Therefore, the most important things first are, ensure the measurements in post 20 are accurate, confirm whether the dynamic deflection in the post 2 video is 264 mm or not, and post dimensioned sketches (side view, front view, etc.). You could also post dimensioned sketches of your proposed redesign.

Last edited: Dec 23, 2013
24. Dec 23, 2013

### 256bits

Two points:
1. check the welds at the bottom where the upright meets the extension on the floor. That is another weak spot that could fail unexpectably. The thing was bouncing around in your test and a weld could have a crack in it.

2. Do the testing in a safer manner. With a failure, you will not be able to move out of the way fast enough to avoid injury. You might think so, but it will fail quicker than you can run.

25. Dec 23, 2013

### Stickyman

I'm 100% the dimensions in post 20 are correct.
I think the fact that it's only loaded in one side causes the frame to twist, there fore adding to the distance between 20kg barbell and 410kg barbell at static measurement.

Here is a very poor sketch, all materials are 100x100x3 RHS or 75x10 flat bar.

Why are you not sold on the catching arms?
It would elevate all of the stress on the bolts, and load the frame symmetrically.
It also shouldn't interfere with the use of the monolift.