Need some advice on winch drum calculations. 1

[matkyne]

I searched the forums for this information and came up empty handed. So if this has been asked and answered before I apologize.

I am working on a project to make a winch lighter. The original design required it to pull over 15,000 lbs, but that load requirement has now been reduced to 9,000 lbs max, working load is 7,500 lbs.

I am trying to evaluate if I can remake the drum out of a lighter material now that the load requirements have been reduced. The winch is designed as a direct pull drum, and in worst case conditions (but likely to never actually be used) the full 9,000 lb load is applied at the all-cable-in condition with 8 layers of cable on the drum.

I used Blodgett's "Design of Weldments" section on drum design to calculate the forces applied to the drum (compressive force normal to the axis) and the outward force applied to the rim from the tension in the cable. I did the math and the vector analysis and the numbers I calculated seemed to be too high.

The reason I think this is because I have read the original FEA report for the drum when the cable load was much higher, but the loads that were applied to the drum were much lower than my calculated values.

So I think there must be something wrong with understanding of the way the loads are applied. The example in the book only consisted of 3 layers of wraps. My drum has 8.

In the example used to calculate the compressive load on the drum, the book reads:

"Although each succeeding layer of the cable should add to the pressure against the drum, the outside layers will tend to force the proceeding layers into a smaller diameter, reducing their tension; in like manner, their pressure against the drum will be reduced. For this reason only the effect of the outer two layers will be considered."

I was unsure how to scale this for my application. As previously stated, his drum only has 3 layers. This can mean that only the two outermost layers on any drum contribute, or that the top 2/3 of the cable layers matter. Any one have any experience with this?

The next question has to do with the outward force that the cable applies to the rim or walls of the drum. The formula given to calculate the inward radial force is:

R_n = (tensile force in the cable in lbs)/(radial distance to application of the force)

Where R_n is the wrap number

You solve this for each wrap of cable. The units for R_n are lbs/in of circumference.

Once you have these values you use vector analysis to find the forces applied to the rim.

In the given example the force vectors line up and are accumulative. But again his drum only has 3 layers so he only considers a maximum of 3 vectors. My assumption is that I would need to consider all 8 layers of the cable for the analysis. Once again I am getting some crazy high numbers here.

I am at a loss here and any help would be appreciated. Sorry for the book, but I wanted to fully state my problem.

 

[som1973]

Hi Matkyne,

According to DNV Standard, the hoop stress must not exceed 85% of the yield stress of the material. Therefore, the thickness of the drum must be sufficient thick to ensure that the drum will not buckle under the wire rope tension.

Minimum Thickness of Drum Required as per DNV Standard =

Rope Factor x (Ultimate Pulling Force x 1.1)/ (0.85x Yield Strength of Drum)x Pitch of the groove on drum.

Consider rope factor = 1.75 for more than one layers.

 

[chicopee]

Did I read this right? eight layers of cable on a winch.
There good reference out there, for one that I know is Shapiro's Cranes and Derricks; Also ANSI/ASME standard on cranes may help you out.

 

[Cockroach]

Yup, the old wireline drum problem! I have worked on this problem about seventeen years ago, crushed eight or nine drum designs, then gave up. It is one of the most difficult problems i'e tried to solve, probably the most frustrating.

You are correct, a said length of wire rope is spooled onto the core of the drum, line tension adds a compressive stree on the drum. This can be calculated fairly accurate as the length of wire rope. The next few wraps are spooled onto the drum, obviously under the same line tension. Doing the math, the new radius of the layer sits in-between the individual wraps below. This is common for all subsequent layers. Known as the Keplar Packing Problem, or simply the Cannon Ball Problem, you can determine the geometry and calculate the new radius for that layer of spooled wire.

But the second and layer numbers thereafter add a lateral force to the outside flanges, hence a bending moment to the drum. This is the trick, depending on the drum core geometry, sometimes the wire rope is in full contact with the flange, other times not, or partially cradled by the layer below. The literature is clear that you reduce layer tension below by half, and add it to the new core pressure by the above layer at that new radius. Plus the bending moment converted back to core pressure induced by the ens of the layer, depending upon contact with the flange. And that depends on drum geometry allowing the rope to be cradled or not, or partially.

And that is the problem I have been working on, intermittently for seventeen years. There is no literature to fully describe the mathematics, particularly if the wire rope spooled to the drum is jarred. That is even a bigger headache.

There are past threads, yes that welding book is probably the best source of information, API specifications help somewhat, also CSA. I'vecome very close in solving this in FEA, but never fully. I do not buy into the mentioned model of forgetting the layers contributing to core crush, as they are relaxed from above compressive forces. Actually flange loading increases and the end OD the drum fails due to bending. Also, the problem is best solved as an iterative solution as I have described above. I leave the details for you to rework based on our dialogue.

Definitely one of the most entertaining and difficult questions in the oilfield industry. Good luck, LOL!

Regards,
Cockroach

 

[GregLocock]

The decrement in load applied to the drum by successive layers depends strongly on the ratio of radial to axial stiffness of the cable, if contact is above a critical friction value. In practice in a real life case with a cable with a high ratio of axial to radial stiffness we measured no increase in drum pressure after the first seven layers at constant tension, whereas when we lubricated the cable we did not find a practical limit.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Cockroach]

Very interesting, Greg. I never modeled friction between contacting layers. Thermal, yes. Wireline drums used in our high Artic can go from -60C to room temperature in a matter of hours, outside ambient to a truck parked in the shop over nite for example. I've had that difficulty including jarring operations or line tension variation during spooking.

Regards,
Cockroach

 

[Cockroach]

It took a little searching, MatKyne, but this is one of the papers I used in my design strategy concerning wireline drums. I trust it will be of use to others as well.

Regards,
Cockroach