Taper Fit Couplings

Consider the semi-angle of tapered part as=α

the friction coefficient at assembly =μ1

the friction coefficient at disassembly =μ2

The assembly force being Fass the disassembly force will be

Fdis= Fass *(μ2*cos α-sin α)/(μ1*cos α -sin α).

It is necessary to use 2 values for the friction for several reasons:

-at assembly the parts slide one over the other so that a "dynamic" friction occurs.

-at disassembly the parts are stuck and the stationary friction has to be considered since they will not move as long as the force is not great enough.

Many times at assembly the parts are lubricated and at disassembly the lubricant was expelled and the surfaces are corroded. Any way when 2 surfaces are pressed on each other for a longer time the roughnesses interlock and increase the starting friction till a sliding occurs.

As you see you should first compute or measure the assembly force.

Are you asking what the hub slip torque is? Or the mounting or unmounting force required? I am not sure what you mean by the "force required to pull apart a properly fit taper coupling."

Kind of both

My architect is concerned that the taper fit does not have enough hold on the shaft to keep it from possibly pulling out from the coupling. This could possibly flood the engine room which has happened before but it is rare, probably a poor taper fit. My argument is based on experience more than engineering know how. I have seen tremendous forces bend shafts,distroy propellers and literally tear the drive gear assembly in two but the coupling was still in place. The fit has to be about 80% or better contact typically blue fitted so the two parts lock togethe once heated and driven or pressed on. he wants to add an extra backup of threads and double nuts to keep the coupling from pulling out and I want to use a more traditional keeper plate with bolts. I know it works I just am trying to support the arguement that the taper properly fit is considerably stronger than the 6" nut he wants to back up the coupling which adds more complexity for installation and removal as well machining.

Thanks

T6he conicity is according to your values 1/12 which leads to an semi-angle of 2.386°. The friction coefficients are usually at least 0.1...0.12 to which friction angles of 5.7...6.84° correspond. So that you are far from self - unlocking.

The only problem is as you mentioned the quality of the fit since the shorter the contact area the smaller the locking force. The design with 15.5" of length for a dia of 10" is correct too. The second aspect is how thick is the wall if the wall is too thin the press-fit is not strong enough. It should be at least 0.3 of diameter. Better more to give a high stiffness in tangential direction and allow high contact pressures.

Which grade do you use for the bolts? How do you control the pre-loading force ?

To compute (estimate) the what I named the disassembly force it will be necessary to have a sketch with the coupling dimensions and the axial shift between load less contact and final position in order to estimate the contact pressure in the fit. If you assembly by pressing then it will be good to know the used force to put the coupling in its final position.

The coupling OD is 17" and is usually heated and drove on with strongbacks and wedges. Some shipyards will use a hydraulic jack with 20 ton rating and as long as the coupling draws up onto the taper approx 1/16-1/8" it will then be finally secured using the bolts and Keeper plate or a large nut. bolts would be grade 8 typ and large nut would be same as shafting a low carbon material equivalant to ABS grade 2 which has less strength than 4140.

Thanks

Ratios Fdis / Fass for different friction coefficients.

µ1

µ2

0.08

0.1

0.12

0.1

0.479

0.412

0.361

0.12

0.644

0.553

0.484

0.14

0.808

0.694

0.608

The interest is to have a small assembly friction <0.12 = good surfaces and lubrication which could be expelled in order to obtain a as high as possible friction at disassembly..

For the 20 Ton assy force the minimal force for disassembly will be 0.479x20=9.6 To

The 4 bolts should bring a pre-load high enough so that in the worse case the coupling will not be subject to an axial force > 9.6 To in disassembly direction.

Since the bolts are very elastic in comparison to the pressed coupling most of any pulling force will affect the press fit.as soon as it is > bolts pre-load. This makes a control of tightening almost compulsory.

The outer dia of coupling is ok since > 16".

Sorry for the presentation: I tried to copy a "word" table but it does not work so you should read the values as follows:

after µ2 the 3 values are the considered µ1 values for the estimations

for every other value between 0.1 and 0.14 considered for µ2 you have the ratios values for Fdis/Fass for the µ2 valeus above the group of the 3 values in thick caracters and for the 3 µ1 values from 0.08 to 0.12.

If you send me an e-mail address i send you the table in xls or word format.

cjtugs@gmail.com

Thanks