Re: U-joint phasing and drive shaft theory (long)

[IHMagellan@domain.elided]

After sleeping on this one, I figured out in the shower this morning why
there's the confusion here.  Joel, you're right that the U-joints need to be
90 degrees out of phase.  This way when the angular velocity is increasing
through one joint, it's decreasing by the same amount through the other joint.
This is assuming that the axes of the transmission output and pinions shaft
are parallel, givng the same bend angle at each joint.  The confusion here
lies in the terminology.  For the joints to be 90 degrees out of phase means
that the input sides of the two joints are 90 degrees apart.  The input on the
transmission u-joint is the obiously the side hooked to the transmission.  The
input side at the pinion end is the half of the u-joint attached to the
driveshaft, itself.  So, if the two inputs are 90 degrees out of phase, this
means that the ends of the driveshaft have to be parallel to each other
because the end at the transmission is an output and the end at the axle is an
input.

As for rotating in circles about the geometric center of the shaft, imagine a
line running from the center of one bearing cup to the center of the opposite
bearing cup (we'll call this L1).  Let's say L1 is connecting the bearing cups
on the transmission side of the transmission u-joint.  Now imagine a second
line (L2) connecting the other two bearing cups in the cross, the ones on the
driveshaft side of the transmission u-joint.  No matter what the anglular
position of the joint, the midpoint of L1 is always connected to the midpoint
of L2.  If the u-joint is centered on the transmission output shaft and on the
drive shaft, then the center axes of those two shafts always intersect at this
same point.  Now let's imagine L3 connecting the bearings on the driveshaft
side of the pinion u-joint and L4 connecting the bearings on the axle side of
that same u-joint.  Using the same argument, the center axis of the driveshaft
intersects the center axis of the pinion shaft at the point where the
midpoints of L3 and L4 intersect.  If the center axis of the driveshaft is
defined by the the intersections L1 with L2 and L3 with L4, then the only ways
for for the shaft's center of mass to move in circles around this theoretical
center axis of rotation is 1) for one or more of the u-joint yokes to be
mounted off-center on it's respective shaft, or 2) for the driveshaft tube to
be off-balance due to variations in wall thickness.  We know that due to
manufacturing tolerances, both are true to some degree.  We attempt to solve
this problem by balancing to move the center of mass back to the theoretical
center axis of rotation.

Sorry to have to temporarily pepetuated the confusion by not catching the
terminology problem before I sent the last post.  Hopefully this one will
clear things up.

Jerry Muncie


In a message dated 98-02-10 21:58:04 EST, you write:

<< Bill is right about the phase.  Additionally, whether the u-joints are in
 phase or 90 degrees out, with a theoretically perfectly balanced driveshaft,
 the center of mass does not go in circles around the geometric center.  And
 don't try to argue that theory and reality aren't related.  You're trying to
 achieve the model by balancing, and putting the u-joints 90 degrees out of
 phase will only add another problem, not solve the original one.
 
 Jerry Muncie
 
 In a message dated 98-02-10 20:57:57 EST, you write:
 
 << Bill,
  
       That's close about the speeding up and slowing down, except 
  for one fact.  If you have the yoke in phase, meaning that if you 
  draw a straight line from the center of one side of one yoke, 
  down the entire shaft and to the center of another side of the 
  opposite end's yoke, this line should not be drawn along the side 
  of the shaft.  If you put a bar in one yoke as if to turn the 
  driveshaft, a bar placed in the other end, should be at 90 
  degrees when viewed down the axis of the shaft.  Thus the 90 
  degree phase.
  
       Anyway, now that you're confused (I hope not though), the 
  key to keeping the drive shaft from vibrating is the center of 
  mass of the shaft.  The axis of the shaft actually moves ever so 
  slightly when at an angle.  The problem is when the u-joints are 
  not phased (bars are parallel when placed throught the yokes) the 
  entire center of mass revolves around a radius that leaves an 
  offset (eccentricity) somewhat deviated from the center line axis 
  of the shaft.  If the yokes are phased, then the shaft ends will 
  oscillate around this eccentricity, BUT, the center of mass, 
  stays at the geometric center.  This allows the ends to cancel 
  eachother out and minimize vibration.  The reason the Double 
  cardon CV joint is out of phase at the ends, and in phase for the 
  shaft itself, is that there is a ball in the center of the CV 
  that keeps the eccentricity lined up with the centerline while 
  rotating.
  
       I wish I had a piece of paper to show this.  If the joints 
  are set a 0deg. phase the drive shaft will actually act most like 
  a jump rope, all the mass revolves around a center line.  Where 
  if the joints are 90 deg phased, then the shaft acts like it's 
  suspended at it's center and the ends make circles, like two 
  cones connected at their vertices.
  
       I really hope this helped, if not, I'll try to figure out 
  another way to explain it.  I played alot with the Chev when I 
  had it.  Since it was a 127 in. Wheelbase, and a mated T/C, the 
  rear shaft was 6'6" long and the front was about 29".  The 
  Travelall with the divorced T/C looks to have almost equal length 
  shafts front and rear.  Also the Chev had a 3" rear shaft, so 
  when it wiggled, you could really feel it!
  
       Thanks for the Trans info, I'm working on a game plan now. 
  
  -Joel Brodsky
  
       '76 Scout II 345/tf727
       '75 T'all 150 4wd 392/tf727
  
  -----INCLUDED MESSAGE FOLLOWS-----
  ------------------------------
  
  Date: Tue, 10 Feb 1998 14:33:36 -0700
  From: "Bill Thebert" <bthebert@domain.elided>
  Subject: Re: U-joint geometry, A/T removal
  
  Joel Brodsky writes:
  
  >      About the phasing, you are correct.  Whoever has rattles, 
  > make sure the yokes on the ends of the drive shaft are 90 deg. 
  > out of phase.  If they are in phase,then the whole shaft 'floats' 
  > and rattles your fillings out. 
  
  Joel:
  
  Perhaps I missed the beginning of this thread, but this is contrary to
  everything I understand about U-joint construction and operation.  
  
  The motion of a bent u-joint is "non-uniform" -- that is, if the "driving"
  shaft is rotating at a perfectly consistent 100 rpm, the "driven" shaft on
  the other side of the bent joint will actually be speeding up & slowing
  down twice per revolution.  Sure, the driven shaft still *averages* 100
  rpm, but the motion is not uniform.
  
  The purpose of the "equal and opposite" u-joint angle rule is to reverse
  this process.  By inputting an irregular, non-uniform motion to another IN
  PHASE u-joint at the very same operating angle, the driven shaft will have
  a resulting *uniform* motion equal to that of the original "driving" shaft
  in the paragraph above.
  
  Putting the joints 90 degrees out of phase will compound (double) the
  non-uniformity of the driveshaft motion rather than "neutralize" it.
  <tranny info snip>
  
  Good luck.
  
  Bill Thebert
  The Binder Bulletin
  http://www.binderbulletin.com  >> >>