The Search for Shangri-La, Nirvana aka "The Perfect Python Pivot" aka the PPP

Joined
May 31, 2013
Messages
3,981
Location
South Benfleet, Essex, England, UK
Dear friends,

As you know, one of our steamed esteemed members has been suffering for many thousands of miles (approaching 6,000 now?) with a weird liking for delta trikes of the genus "Pythonidae ".

We have tried counselling, group-therapy and "interventions", we even hired an expert in de-programming, but all to no avail. He won't budge, the obsession is too strong.

Given that we are not going to change him any time soon (the old fellow is far too set in his ways now) a concerted effort has been made to help him cope with the trials and tribulations of this affliction.

There are lots of ways to joint a trike of the genus "Pythonidae " and some are better than others (allegedly) and some are less robust and more prone to seizures and failures.
These range from:

  • Headset bearings.
  • Rod-end (Heim Joint/Rose-Joint) bearings.
  • Angular Contact bearings.
  • Tapered Roller Bearings.
Each of these approaches has been done and each has its relative merits in terms of:

  • How easy it is to deploy.
  • How much it costs to implement.
  • How robust it is in service.
  • How much maintenance it requires to keep performing optimally.
It will come as no surprise to anyone that the mid-point solution in terms of robustness, ease of deployment and moderate cost is the Rod-end/rose-Joint/Heim-Joint (call it what you will) solution.

However it is not without its drawbacks too.

  • It can have a short life before its friction and resistance to angular change begins to increase.
  • It can be difficult to align properly.
  • I can be difficult if not impossible to lubricate.
All that being said, our friend has managed over 6,000 miles on the snake-bike quite happily (albeit with regular rod-end changes).

This got me wondering how difficult/costly would it really be to implement the “Rolls-Royce”/Other-end-of-the-scale solution? In my eyes this would be the “Tapered Roller Bearing”.

Now I am no “Sandman” and I do not have that background in engineering practice nor do I have access to a whole host of quality machinery; but I do like to tinker with things and I love a challenge. I also have a Delta trike that has all the elements of a MBB style bike that I have found difficult to get along with and a conversion to a mid-jointed trike of the Pythonidae family is an interesting proposition.

I did some web-searches and found some modestly priced Tapered Roller Bearings that were rated for 9,500 RPM and >20kN loads (4,000) pound loads. Now the pivot we want is not spinning through 360 degrees and the loads are certainly less than 4,000 pounds, so I think they will do an admirable job.

So I set about putting together the elements that might provide the “PPP” (Perfect Python Pivot). I bought the following items to make the pivot itself from:

  • A pair of budget Tapered roller bearings of modest size (40 x 17 x 13m) @ £4.08 each.
  • A 300mm section of 45mm seamless cold drawn tube with a wall thickness of 2.5mm to provide the outer housing for the 40mm bearing races @ £10.67
  • A 300mm section of 20mm seamless cold drawn mild steel tube with a wall thickness of 4mm to provide the housing for the bearing inner races and the 12mm axle-bolt @ £8.99.
  • 0.5mm Shims for use in adjusting the end-float of the PPP assembly. (qty 50 for £15.98).
  • Other tubing pieces suitable for use as inserts inside the outer tube to keep the outer bearing races in position. I had some suitable tubing that was slightly too big so I slit it and took a small slice out. This gets puddle-welded inside the main outer PPP tube.
Design of the PPP.

Ok, so I have never done this type of thing before so I may have made some errors but I tried to keep the thing as simple as possible (though you may indeed think otherwise).
The assembly consists of the bearings and two main sleeves which are used to position the bearing inner and outer pairs in the correct relationships and revolve around a 12mm central axle/bolt
The bit I could not “fudge” my way around or “hack” was the central sleeve to support the pair of “inner races” of the bearings.

So, I asked around and got a chap to turn some for me from my tubing, according to a drawing I provided (see 3D representation below).
Sandman (John) was always badgering me to learn how to use Solidworks CAD so I have (a tiny bit).


The chap had an hourly-rate and so I got my tube “turned” [pun] into 4 parts in that 1-hours cost. So I can create 4 PPP’s should I need to. Cost £50. So £12.50 each.
It has made me think longingly about buying a small hobby-lathe though.
These are really dangerous thoughts and I must resist at all cost$$.

Putting it all together.
The hip bone’s connected to the leg-bone……

In the diagram below nothing is to scale but the main elements of the assembly are shown and their relationships to each other. The PPP is bolted through two end cheek-plates and is free to rotate around the axis of an M12 pivot bolt that runs through its long-axis. The pivot bolt and washers hold the two inner races of the Tapered Roller Bearings (TRB’s) firmly into the ends of the PPP outer tube and against the stop-shoulders of the inner PPP sleeve tube and places the rollers into contact with the TRB outer races which are inserted into the PPP outer tube and held in place by a welded in sleeve that prevents its movement within the PPP outer tube.



The diagram below shows an exploded simplistic view of the PPP alone.


The following pictures show the actual parts of the PPP as they really are and how they fit together to form the pivot.

The Inner Sleeve.
The inner sleeve is a machined element made from cold drawn precision tube 20mm in diameter with a 4mm wall thickness. To keep the “PPP” at a modest and not ungainly size an overall length of around 68mm was desired. The two ends of the inner sleeve were turned down to 17mm OD for a distance of around 13.5mm. These reduced ends are to be inserted into the “inner-races” of the Tapered Roller Bearings and sit on the shoulder of the turning.



The outer races of the Tapered Roller Bearings are pre-machined “cups” that must be held in the correct position in relation to the inner race with the rollers. Effectively, the base of the metal cup should be 13.25mm from the front (outer) face of the inner race. If you can get your head around this in one go, you are a far better man than I am. It is dimension “T” in the diagram below.



This means that as the only fixed item in our catalogue of parts is the central machine sleeve then we must place the inner bearing races on the inner sleeve and measure the distance across the outer ends of the races in this position (see below).

The machined cups that form the outer races and that are held in the outer tube must therefore have their inner edges 66.29 (the distance between the shoulders of our inner tube) - 26.5mm = 39.79mm apart. This will be the desired length of the outer tubes bearing stop sleeve.

In my first attempt to make a “PPP”I tried to accomplish this separation of the outer race cups in the outer tube with 2 separate stop sleeves. While this might just possibly be easier to implement it has some drawbacks.

  • It places the required puddle-welds to hold the stop sleeves in place too close to the ends of the main outer tube and the outer bearing cups and leads to distortion of the tube and the cups are then more difficult to remove & insert.
  • You now have a minimum of 4 puddle welds whereas with a single tube you only need 2, meaning more heat, more distortion.
So, I junked those parts (outer PPP tube and the two welded-in stop sleeves) and started afresh.

I cut two new tube sections for the outer PPP tube and its inner sleeve.

I don’t have precision equipment, just a chop-grinder) so I cut these tubes over-size and filed them back to exact size and squared the ends as I went.

The picture below is of the outer race stop sleeve. It is in fact a 43mm tube section and is too big to fit inside the 40mm ID of the PPP outer tube. So a slice is cut out and it is re-rounded with a vise and held closed with a jubilee-clip as it is inserted into the PPP outer tube.


Here are all the parts that will be placed inside the Outer PPP tube, all stacked together.


Below is the stack of internal components, and below it the outer PPP tube into which they will all be placed.



Then it was time for a dry-run with it all together and a 12mm bolt through it. Would it spin freely?
Yes, it was all good. So the PPP “inner” tube and the Tapered roller bearing inner races were removed and the PPP outer tube drilled through to puddle-weld the outer race stop-sleeve in place.
After this would it still turn ok?
It was all re-assembled and bolted up and yes, it’s just fine & dandy.
So the quest for the holy-grail “Perfect Python Pivot” is over.

A kit of PPP parts will accompany me to the BHPC rally on the August bank holiday where they shall be handed over to our dear friend to aid him in his affliction ;-).
 
Last edited:

Radical Brad

Garage Hacker!
Staff member
Joined
Mar 13, 1999
Messages
6,078
Location
Kakabeka Falls, Ontario, Canada
This looks like a great solution as well as the perfect job for tapered roller bearings due to the combination of radial and axial loading.
I look forward to seeing the result slither down the road for a test run and ride report.
If you send me your basic model, it would fun to do a stress simulation on it. I use SW2019.

Brad
 
Joined
Feb 7, 2008
Messages
4,575
Location
Nottinghamshire England
I almost understood some of it :)

I will have to fight my natural tendency to throw it on the workbench amongst the other stuff ' that would be good to use when I get around to it ' ;)

all the best Paul
 
Joined
May 31, 2013
Messages
3,981
Location
South Benfleet, Essex, England, UK
This looks like a great solution as well as the perfect job for tapered roller bearings due to the combination of radial and axial loading.
I look forward to seeing the result slither down the road for a test run and ride report.
If you send me your basic model, it would fun to do a stress simulation on it. I use SW2019.

Brad
I would love to Brad, but the only part I drew was the central sleeve-tube and that took me ages (Novice-R-US). :ROFLMAO:
 
Joined
Oct 19, 2012
Messages
2,384
Location
Wakefield, UK
Nice engineering there. I used a bottom bracket on mine rather than the headset as it looks designed to take a bigger load than the headset yet also comes free with a donor. Yours will look so much neater and professional though.
 
Joined
May 31, 2013
Messages
3,981
Location
South Benfleet, Essex, England, UK
Nice engineering there. I used a bottom bracket on mine rather than the headset as it looks designed to take a bigger load than the headset yet also comes free with a donor. Yours will look so much neater and professional though.
Thank you, I am hoping it will be nigh-on unbreakable, it certainly weighs enough (462g), although I have not weighed a BB Shell with a cartridge in it for comparison.
Time will tell.
 
Joined
May 31, 2013
Messages
3,981
Location
South Benfleet, Essex, England, UK
An update.
The PPP itself is more or less complete now. All that it lacks is an M12 bolt of the requisite length and a Python to mount the PPP base-plate to.
It is the manufacture of the base-plate that I have been undertaking for the past day or two.
The baseplate is made from 4mm cold-rolled mild steel plate. It has two captive M8 bolts embedded within the base-plate so that the whole assembly may be attached to any bike frame once the required sleeved holes have been created in the recipient frame. In the pictures below you can see the finished base-plate with the PPP mounted within it. The temporary pivot-bolt is obviously over-length and is merely being used as a temporary measure.





And here is a little video of it spinning and the assembly of the PPP into the base-plate (a job one would do only when re-greasing or replacing(unlikely) the bearings).


That's all folks.
 
Joined
Oct 19, 2012
Messages
2,384
Location
Wakefield, UK
That baseplate looks like it will hold the weight of an elephant. Given it'll likely be bolted to a substantial piece of chassis anyway cutting it into a two piecer could save you a chunk of weight with the bit between the ends currently duplicating the job the chassis member could do. You could then weld the ends to that chassis member saving the weight of the bolts and sleeves too.
 
Joined
May 31, 2013
Messages
3,981
Location
South Benfleet, Essex, England, UK
That baseplate looks like it will hold the weight of an elephant.
You are right! I will be riding the trike, so it will have to hold the weight of an Elephant! :ROFLMAO:

However, that central bit of the base-plate weighs only 83g :eek: so I think on balance it is not worth removing and it makes the sub-assembly extremely rigid and "bomb-proof".
 
Joined
Jun 22, 2015
Messages
127
Location
Ontario
Interesting project. All of the trailers that I have owned use opposed taper roller bearings for the axles. The bearings are not "preloaded", rather the axle nuts are fastened finger tight (after torquing is done to seat the bearings, and then untorqued) and locked with a split pin or a retaining ring on the nearest flat. I presume that your assembly just uses shims as required to take up any bearing play.
 
Joined
May 31, 2013
Messages
3,981
Location
South Benfleet, Essex, England, UK
Interesting project. All of the trailers that I have owned use opposed taper roller bearings for the axles. The bearings are not "preloaded", rather the axle nuts are fastened finger tight (after torquing is done to seat the bearings, and then untorqued) and locked with a split pin or a retaining ring on the nearest flat. I presume that your assembly just uses shims as required to take up any bearing play.
I too understand that that is the standard process for the use of TRB's on stub-axles.
In this case too the TRB's are in an opposing pair and the distances between inner & outer races are controlled by race separation using the inner/outer sleeve separators. Shims on the inner sleeve are used to add slack. I was lucky in that I very, very carefully set out the inner to outer race spacing and I feel it does not need any shims. Each of the 2 TRB's are rated for 9,000RPM and >20kN loading. The use I am putting them too requires <90-degrees of rotation (Zero RPM).
I think it will be OK (aka famous last words), but if Rose-joint bearings can be utilised for this purpose and not catastrophically fail, then think TRB's will be even better. ;)
 
Joined
May 31, 2013
Messages
3,981
Location
South Benfleet, Essex, England, UK
I was inserting the crush-tubes in a section of frame when it all went horribly wrong. :eek:

“The best made mice of men with plans do oft-times FUBAR-up” – Bill Wag-a-dagger.

I thought I would ensure perfect alignment of my crush-tubes with the welded in M8 studs on the base plate by tacking the tube in place while the studs were inserted through them.
The plan/intention was to then remove the PPP base-plate and studs and complete the welding in of the crush-tubes.

Unfortunately, the tacks went rogue and managed to melt into the studs and there I was with a welded-in part where it should not be.
Cue “head scratching” and that “Uh-Oh” sinking feeling you get when you envisage hours of work going sideways and the possibility of having to start from scratch.

I cut the worst affected stud off with a hacksaw at its base (where it exited the base plate) and attacked the other (less-affected) with a file.
So, one stud amputated and the other now able to be wound out of it's tacked in crush-tube. Phew!

But how to repair the situation?

I removed the PPP out of the base-plate/carrier and centre-popped the hacksaw'd end of the stud. Put it in my pillar drill's X/Y vise and lined it up and drilled it out (3mm pilot then an 8mm hole) for a new bolt.

I finished off the welding in of the crush-tubes, cleaned up the end-faces, filed them back to perfect roundness and tested that an M8 bolt could slip through.

So now I have one end of the pivot secured by a fixed-stud, and the other end by a removable bolt.

Just a temporary bolt for now to test the fixing/mounting.

All good, I am somewhat relieved.



 
Joined
May 31, 2013
Messages
3,981
Location
South Benfleet, Essex, England, UK
As an alternative to the Tapered Roller Bearing PPP, I have also made one utilising Angular Contact Bearings.
Strangely the Angular Contact bearings were MORE EXPENSIVE than the TRB's at £6.38 ea.

I have made this to fit inside the same carrier/base-plate as the TRB version.
This one required no special machining of parts, just carefully matched inner and outer sleeve bearing-race separators.
It took me about 2 hours from start to finish to make this pivot. It is smaller in diameter than the TRB variant (35mm Vs. 45mm) and is lighter (374g).

 
Joined
May 31, 2013
Messages
3,981
Location
South Benfleet, Essex, England, UK
The PPP is getting a revisit for reasons best left well alone as I am in enough trouble for not getting any miles in already.

However, there is a fiendishly cunning Baldric plan to resurrect the delta project which stalled at the "swingerometer of ankle whacking skin-removal" stage with the MBB front end.

Of the two PPP options created one is a bit of a monster with tapered roller bearings (see pages above) and the other was its little brother which incorporated Angular contact bearings in a smaller and lighter shell.

I had brilliantly sensibly made both bearing housings as units of the same length and so they can fit into a common carrier which bolts onto any frame with 2 M8 bolts.

In a further flash of stupidity brilliance I decided that the rotating part of the PPP should itself offer up an attachment plate that any front triangle could be bolted onto and this makes the whole PPP affair very flexible/transferable between projects.

So today I had a play at making this attachment plate. What a palaver! :eek:

I may just scrap this "first-pass" attempt (made from 2 pieces of angle iron welded together and try again with sections of 3mm plate instead. Things always go awry when you try and get too fancy.
What you can see below is one of the sections of angle-iron already trimmed, the other yet to be trimmed and the PPP bearing assembly inserted to show where it fits. The axle of the pivot is 12mm the angle-irons 3mm thick.

I am hoping that the 3mm steel will be robust enough to resist bending and distortion when loaded.





 
Joined
Oct 19, 2012
Messages
2,384
Location
Wakefield, UK
Mounting plates are going to add their weight to any bike / trike plus the need to match that mount on the chassis rather than simply welding the chassis beam to the shell. I can see that it would make it easily swappable but are you going to have that many bikes / trikes and only one pivot?
 
Joined
May 31, 2013
Messages
3,981
Location
South Benfleet, Essex, England, UK
Mounting plates are going to add their weight to any bike / trike plus the need to match that mount on the chassis rather than simply welding the chassis beam to the shell. I can see that it would make it easily swappable but are you going to have that many bikes / trikes and only one pivot?
Completely agree with the weight-conscious assessment. It will undoubtedly be heavier.
But, while I am faffing around making Mk1..99 front-ends and different fixed/suspended/tilting back ends until I get to what I am happy with, it cuts out the heartbreak of chop/weld/chop/weld onto the core bits in the middle. If I make the south side of the pivot 2 x M8 bolts at "X" centres as a standard fixing and the north side to any triangle a 2 x M8 bolts on a 3 or 4mm plate of 75mm x 50mm @ "Y" centres then I have a replicable interface to work to. Seemed to me to be a fair trade-off for 4 bolts and <0.5Kg of plate material while tinkering with concepts and ideas that are at the moment alien to me..

Clearly, once a preferred back-end and triangle layout/design is arrived at, a permanent version without the weight or the changeability can be made.
 
Joined
Feb 7, 2008
Messages
4,575
Location
Nottinghamshire England
I can see that it would make it easily swappable but are you going to have that many bikes / trikes and only one pivot?
The lad has seen the light ' oh brother '

He wants to enter the wonderful world of Delta trikes ?

Now Tadpole trikes are relatively staid ? and very few make you go WOW or ' how the hell does that work ' ?

However with Delta's the front could be :-

Twisting chain
MBB
RWD
Python
etc

The Delta rear could be :-

Straight dead axle
RWD
Tilting
Suspension
any combination of the above ?

So the need for flexibility whilst exploring this brave new world makes sense ?

Same reason I am building ' The Mule '
 
Joined
Feb 20, 2013
Messages
869
Location
Axedale, Victoria, Australia
Website
axerail.coffeecup.com
Late Edition: Such a universal front end or rear end connection could be provided by making it a wall-mounted bracket to provide a very stable, non-existent other end for what you might like to prototype. When you are finished, use it for prototyping the other end.

Single wheel trike ends have a habit of wanting to lay down until their other half is attached.
 
Top