Picking up last week’s traffic light with the two things I left out: the weight of the beam itself, and the bolted connection that started the whole thing.

The weight of the beam
The arm is a hollow tapered tube, so its mass needs an integral rather than a guess.

| Beam mass | 254 kg (2540 N) |
| Centroid | 4.79 m from the root |
| Reaction moment, with beam weight | 12,166 N·m |
Adding the beam’s own weight multiplies the moment at the connection by roughly six. That is the load the four bolts actually have to carry.
What the bolts see

Two separate load paths:
- Shear is carried equally: 2807 N ÷ 4 = 700 N per bolt
- The moment is not. Treating the flange as tilting about its bottom edge, each bolt’s force is proportional to its distance from that edge, so the outer bolts do most of the work.

| Bolt force, far row | 26,285 N |
| Bolt force, near row | 2,920 N |
| Maximum principal stress | 23 MPa |
| Maximum shear stress | 11.5 MPa |
Each bolt is in combined tension and shear, so the stresses come off Mohr’s circle rather than either load alone.
How much preload is needed?

The clamping force has to do two jobs: friction against the shear load, and preload against the tensile load from the moment.
| Clamping force for friction (μ = 0.2) | 3,500 N |
| Preload against the moment | 26,285 N |
| Total preload required | 29,785 N (160 MPa) |
| Bolt proof strength (1.5” grade 8) | ~1000 MPa |
The conclusion
So even with the required preload, the bolts are only experiencing about a tenth of their proof strength.
Bolts in permanent connections are typically preloaded to around 90% of proof strength, which would give this particular connection a safety factor of about 9. That doesn’t seem unreasonable, considering you’d want to replace these bolts perhaps once every 10–20 years rather than routinely.
There is still plenty more to consider when designing bolted joints — the stiffness of the joint and the fatigue life of the bolt among them — but those will have to wait.