Stress Testing Olympic Barbells

With the Paris 2024 Olympic Games right around the corner, I wanted to look at one of my favorite events at the games, Weightlifting. While the incredible strength of the athletes is awe-inspiring and entertaining to watch, from an engineering perspective I found myself internally pondering, “With new records being set at seemingly every Olympics, I wonder how much weight it would take to make the barbell fail mid lift”. Since I have access to Altair Engineering solutions, I decided to find an answer.

The Set Up

According to the IWF (International Weightlifting Federation), the barbell for the official Olympic Games is 2.2 m (86.61 in), a diameter of 28 mm, made of Steel and weighs 20 kg (44 lbs.). Included are collars and sleeves on each side of the bar with bushings and bearings. Since I am not looking to run a nonlinear analysis, nor am I looking to simulate the time steps of a lift, I decided to use Altair Inspire to help answer my question.

The event with the most weight added to the bar typically is the Clean and Jerk. The Clean portion of the motion involves lifting the weight from the ground to your shoulders, then the Jerk motion involves the athlete lifting the weight above their head, locking out their arms. The current Olympic record for weight on a bar during a successful lift was set by Lasha Talakhadze from Georgia with a Clean and Jerk of 265 kg (584 lbs.) at the Tokyo 2020 Games. Since the bar didn’t fail during his record-breaking attempt, I decided to use that as my starting point.

Running the Analysis

Using Inspire, I can use the Material Library properties to check the Yield Stress for Steel (AISI 4142) is around 84.8 ksi. Note, Yield Stress is the point where a material starts to cease elastic deformation and begins plastic deformation. Specifically, this is the point where a material doesn’t return to its original shape after removing the load.

Fig 1: Material Properties including Yield Stress

For this analysis, I’ll start by applying a static load of 100 lbs. of force per plate that’s added. I will apply two constraints on the bar to indicate where the hands are located in a lift. Many weightlifters lift the bar with a wider grip, closer to the sleeves, which can change the results of the simulation, but for this example I went with a standard grip width with each constraint near the knurl markings on the bar.

Fig 2: Constraint locations on barbell

Another variable that can impact the results is the bounce off the chest/shoulders that happens during the clean and jerk motion can add additional stress to the bar, but for this example, I was looking at the basic static simulation of how much of a load can the barbell handle before hitting critical deformation and yield points.

For my first simulation I put 3-100 lb. plates, on each side of the bar, 600 lbs. or 272 kg. As you can see from the results, with this load I am getting some significant stress of just over 61 ksi. In the report and visualization options inside of Altair Inspire, I also have the option to check the Percent of Yield, which in this simulation was 72.66% of yield.

Fig 3: Stress and Yield result from initial analysis

Since my first simulation failed to hit the yield point for my bar, I upped the ante by adding an additional 100 lb. plate on each side of the bar. The total load at this point is 800 lbs. or 362 kg.

Fig 4: Stress and Yield results from second analysis

I can see based on the results of this simulation that I am very near the yield point for my bar at 82 ksi, or 96.89% yield. I know I’m close. I don’t need to add a lot more weight to answer my question. Instead of adding two more 100 lb. plates, I added one additional 25 lb. plate to both sides of my bar, shown in gray below. This puts the total load at 850 lbs. or 385 kg.

Fig 5: Barbell with additional 25 lb. plate on each side

With Inspire, adding two additional plates was as simple as copying and pasting an existing plate (Ctrl + C and Ctrl + V), then adding a load of 25 lbs. of force to those plates. Users can also right click and Copy/Paste as well. Then I can run the simulation with the same settings as before.

Fig 6: Stress and Yield results of third analysis

We did it! By looking at these results, 850 lbs. of load split evenly on each side of the bar hit nearly 103% of yield on the bar. Now, it should be noted that yield doesn’t mean the bar experiences catastrophic failure, this just means the bar won’t return to its original shape after the load is applied.

Just for fun, I am going to see what would happen if I maxed out the bar and put 6-100 lb. plates on each side of my bar, 1,200 lbs. or 544 kg.  

Fig 7: Stress and Yield results from final analysis

Adding that extra weight significantly impacted the results, as expected. The stress kicked up to 23 ksi, and Yield of 145%. Remember once I enter the plastic regime (past the yield point), I need to run a Nonlinear study, since the stress results I am capturing here are for a linear static study. With Altair solutions I can run a nonlinear study to capture the plastic deformation in Altair HyperMesh or Altair SimLab.

One additional note is that with running 4 separate simulations, none took longer than 3 minutes and 26 seconds to run using the built-in OptiStruct solver, which I have a standard setting of using 2 CPU cores. I feel satisfied that I have a rough answer to the question of how much weight it would take to make the barbell yield. Now, I look forward to seeing how long it will take for human performance to reach that level. If you have any questions on Altair Inspire, let us know!