The weight of visible evidence

A recent technical query from a member company involved a 35 tonne double girder overhead crane on a gantry of long spans supported by knee brackets from a portal frame building. The gantry was new and had apparently been designed specifically for the building and the crane. However, as the crane travelled down the gantry, it was clear that there was a problem with the span. The wheel flanges were binding on the rail, causing the gantry and building to deflect. Moreover, when the crane was positioned at the centre of a gantry span and the crab was traversed across the crane bridge, the gantry beam visibly deflected sideways just from the dynamic forces arising from the unladen crab's own weight. Under full load, the possible consequences were obviously even more serious, particularly if the driver was a little clumsy.

Clearly something was wrong – even though the contractor who had manufactured and installed the gantry claimed to have checked the calculations and found them to be correct. Other than to confirm our member's own judgement and suggest a course of action, there was little we could do to resolve this particular problem in the short term. However it did highlight a number of issues which can and often do arise with crane gantries supported from the building structure.

The first of these is control of the span. The building involved in this particular query was a portal frame design. This is a modern and efficient means of producing a wide clear span. However, the effects of external influences such as wind and snow, together with internal loads such as services, water tanks and process machinery, can all cause the frame to flex at different points according to circumstances. As a result, the span can vary significantly from day to day, often beyond what is safe for a crane gantry. If headroom permits, the usual way of addressing this problem is to add ties across the portal frames to limit the movement.

Another issue to consider is the alignment of the gantry. There are several alignments to check, and all are measured when the gantry is not supporting the crane. The maximum tolerance on the span varies according to the length of the span involved. For cranes up to 15m span, the maximum is 3mm. For crane spans over 15m, the maximum tolerance is 3+0.25(S-15)mm up to a maximum of 15mm, ie an additional 0.25mm per metre of span over 15m. Each rail must align both horizontally and vertically to a straight line mean datum. The maximum deviation in each plane from this datum should be 10mm, with any local deviation not exceeding L/2000 measured over a length L of not less than 2m. Finally, the difference in height between the two rails should not exceed 0.001% of the span, with a maximum of 10mm.

In order to align the gantry within these tolerances, the gantry beams and rails must themselves be sufficiently straight. This is more likely to be a problem with long spans, as in the aforementioned query. Also, the design of the supports must allow for both horizontal and vertical adjustment. This may sound obvious, but I well recall seeing a new gantry connected to the portal frame by short knee brackets bolted to the web of the gantry beams, a design which renders alignment almost impossible to achieve.

The design of the gantry must allow for the dynamic forces arising from longitudinal and transverse movement as well as vertical lifting. This is something not always appreciated by general structural designers. Also, depending upon the age of the crane and the standard it was built to, there may not be any built-in protection against either collision of the crab against the bridge end stops or the entire crane against the gantry end stops. In both cases even low speed collisions can produce very high shock loads.

Having ensured that the design takes all these factors into account and that the gantry has been properly aligned, there remains the matter of testing. The simple operational test of travelling the unladen crane down the gantry will double check the span alignment. Measuring deflections of the structure under load and comparing them with those predicted by the calculations should verify the calculations. Observing how the travel motions of the crane perform under load will also offer an opportunity to confirm the height and level alignments of the gantry; any difference in speed between one direction and another will indicate the existence of a slope.

Calculations are an invaluable tool for designers, but only if the correct assumptions, data etc are used. Where there is any uncertainty, practical tests and simple common sense observations should always be employed to double check and verify that the results predicted are achieved in practice.