Proper Powr-lok Assembly?

Rebuilding the rear diff on my 68 CJ5 with original Dana 44 and Powr-Lok and ran into a discrepancy in the FSM. It states in the text that "Powr-Lok clutch plates are always assembled with the CONVEX side of the curved plate toward the case." Then what does the picture show? The CONCAVE side toward the case. Nice.

From the FSM (number "1" is the Case, "2" is the dished plate)...

When I pulled it apart, the convex side was indeed toward the case (as per the FSM text). To make matters worse, I found a PDF for Dana 60 and 70 Powr-Loks and it said to make sure that the plates were reinstalled in the same orientation as they were removed, but then the diagram in the PDF shows the concave side toward the case. Asked my engine rebuilder, he thinks convex toward the case (the way it was originally assembled) is correct...which would make the diagrams from two different sources wrong.

Just wondering if anyone has a more definitive source of info on this. I have it assembled convex side toward the cases currently, but can easily change it prior to installing it in the axle housing if needed.

Edit: Leaning toward CONCAVE side toward the cases due to the below diagram from the Dana 60/70 PDF as the only way you would get "typical wear" on the inner edge of the plates (near the splines/hub) is if the concave side is toward the cases. Obviously this would mean my original assembly, the FSM text and my engine builder are all incorrect.

 

Thanks Howard. This service bulletin also clearly states in the text, "The dished plates in the disk and plate set are always assembled with the convex side toward the case." (likely copied straight from the FSM). It also references Figure 3 if using new disk set (which I am), but Figure 3 is not clear enough to tell which way the "Domed" plate is facing.

So at present, I have two sources of text that say convex side toward the cases and two sources that show a diagram with concave toward the cases. Tie-breaker anyone?

 

In truth Dana Spicer gives different clutch plate stacking procedures for each model of axle assembly.
Furthermore those specifications have changed over the years.
Page 9 of the 1967 service bulletin says it all.
The specific plate arrangement was superseded over the years.

But in the end I agree with Howard it's not really a big deal.
Make sure that any worn plate faces are not further subject to additional wear.
Occasionally worn belleville plate faces can be mated against the case where they do not rotate
If you have any worn faces that rotate you need consider changing the plates and disks.

 

Thank you gents. Kinda glad to hear this to avoid reworking it. Fortunately, I installed a whole new set of plates. Wear on my originals was very minimal and I have it assembled the same way, so it should be good.

Thanks again.

 

Installing them with the concave side toward the the case will make it easier to align everything on assembly as it pushes the splined clutch discs onto the splined hub (sometimes called a clutch ring). Ed Holbrook (Holbrook Spring Conversion) taught me that trick. Since he's been working on these since he worked at a Studebaker dealership, and has been in the 4x4 business since the '70's, I pay attention to his thoughts.

If you have more than one Belleville spring for each side (the dished clutch disc) you can tune how aggressive the Powerlock is by changing the stacking of the discs and Belleville spring.