The results confirm that for a single notional lane loaded, the Normal vehicle (NA) loading the TMH7 underestimates bending moment effects for actual traffic movements. It is also shown that 24 units of abnormal (NB) loading does not adequately compensate for this deficiency either, NB36 does not share this deficiency. We can also see that the Eurococdes load Model 1 over estimates actual traffic loading.

The following figure shows that although at SLS the NA loading cannot effectively estimate actual traffic loads, the safety factored applied at ULS produces bending effect results that ensure structural safety (only marginally for spans ranging 5m).

Already after two notional lanes loaded, the NA ant LM1 are beginning to parallel one another.

It is observed that NB36 after 3 notional lanes are loaded, becomes insignificant compared to those effects generated by type NA loading.

we can now see that after 4 notional lanes are loaded, NA loading and LM1 produce similar results i.e. NA and LM1 are beginning to converge.

For a single notional lane loaded, the results confirm that NA loading under estimates actual traffic loading for spans up to 10m, however as noted in the bending moment results, the application of the safety of factor at ULS compensates for this deficiency.

NB24, NB36 and Eurocode LM1 all produce shear force effects that over estimate actual traffic data, with the Eurocode producing excessively high results. Showing that the Eurocodes need to first be calibrated as they do not model actual traffic data in south Africa effectively either.

NB36 is still producing higher shears than NA loading for spans just over 30m. It can also be seen that the Eurocode produces much higher effect that NA and NB loading, however, we see that NA loading is beginning to converge on LM1.

We can now see that after 4 notional lanes loaded, NA and LM1 are producing similar shears.