If you know the power input to the screw, angle of the blade at the periphery, etc. I guess it would not be too difficult to come up with a theretical assessment of the split of power between radial and axial forces.

An alternative approach would be to design for the full motor power converted to an axial load on the bearing. If you can do that you should be OK for events such as a conveyor jamming due to overload or similar. If this approach comes up with a ridiculously large bearing you may want to do the more rigorous analysis but on a cantilever screw I would have expected the design to be dominated by radial loads anyway.

As far as radial loads are concerned, they are material dependent. Run the conveyor empty and you are into simple bending theory due to the self weight. When you put material into a conveyor thpugh, the material at the bottom can support some of the weight of the flight and reduce the radial load on the bearings. However magnitude of any reduction in effective weight is material dependent. It would be very unusual if this resulted in an upward force greater than the weight of the flight though.

If you are doing bearing calcs, as the conveyor will probably run empty for some parts of its life, working on the self weight of the flight (assuming your material doesn't build up on the flight and increase its weight?) would look sensible.

If you are doing a fatigue analysis on the shaft and are interested in the loads over different periods of the operating life, the only thing I could suggest is to use a test machine, measure the sag when empty and when operating with powder. From that, yopu can back calculate any support which the powder gives the flight. ■

As you can probably guess, I normally take the simple conservative approach and only analyse more rigorously if the simple approach gives an answer which looks too "over the top".

If you can calculae the torque and the angle of the blade at its periphery (which is dependent upon the pitch and diameter of the flight). I would assume that the material directly in from of the blade was acting as a row of ball bearings that are fixed in position. As you suggest it is then a simple force vector calculation to convert that into an axial load.

If that results in a ridiculously large bearing, you may have to do a more rigourous analysis, first try assuming a coefficient of friction between blade and powder.

You could also try bringing in soil mechanics type active / passive pressure theory. Assume the torque resists an active pressure on the powder in the but the pressure imposed back at 90 degrees to that is an active pressure and is consequently significantly less.

I am sure that you realise that the above is meant to help with your thought processes to come to an answer, I am not suggesting it is a standard method of analysis.

One further thought this axial force would be much more of an issue with screw dewatering presses and similar devices. If you don't llike the answer you get from your calcs, try having a chat to manufacturers of this type of equipment to see what they do. ■

Lynn has answered your first question, answering your final two points:

The effect of the flighting on the overall structure is surprisingly low when you do the analysis. I remember checking it years ago and the difference it made was so small it just wasn't worth including in the the overall stiffness calculations.

Creep isn't something I have studied before on a conveyor, I wouldn't have expected you to have a problem with the flight tube though as long as it is rotating. Any bending stressses in one direction would be reversed when it is rotated through 180 degrees.

I am sure there will be some parts which are subject to a load only in a single direction, the weld of flight to tube for example but i would have expected you to have to get to a very high temperature or to have the conveyor static at temperature for a long time before you saw any real problems with creep. ■