Caterham A Frames - check them as part of your annual maintenance

[IamScotticus]

A frame weaknesses have always been the bain of the Seven.  I'd say find a talented welder, give him the original A frame to fab up a jig and he build a solid A frame replacement on it.  If you order an A frame from Cat, or Arch, you'll probably end up changing the radius arms as well due to a length difference. 

Edited July 29 by IamScotticus

[MV8]

I hope the fix was thicker wall tubing or better yet, a thick wall insert on each side so the weight is minimal and away from the swinging-unsprung end. Failure on the first page was not the joints. My splice would work as long as it clears.

 

 

[KnifeySpoony]

In addition to localizing the DeDion tube fore/aft (where rigidity and strength is welcome), doesn't the A frame need to twist as the car rolls? Where the A frame meets the tube, there is no spherical joint- it's rigidly mounted. So it seems the A frame is flexing and thus adding to roll stiffness of the car. Correct me if I'm wrong.

[MV8]

The axle should roll on the A frame center bush. All the axle attachments and are (relatively) compliant bushings but the center would take the most abuse. A spherical center, rigid delrin insert A frame bushes, and sphericals on both ends of the control links should tighten the car and greatly reduce/essentially eliminate any easily measurable deflection of the standard A frame from bind in the original bushes. Just switching to a spherical center should eliminate A frame failures from binding. One reason oem vehicles have such large compliant bushes is to allow movement that could break parts.

 

Impact damage that could easily oval a thin walled tube greatly reduces the rigidty of the tube; likely in a small area. If it is going to flex from binding, the rigidity should be as uniform as practical to spread the flex across as much tube length as practical.

[IamScotticus]

Like a Panhard rod, or Watts link system,  the A frame's porpoise 🐬 is only to restrict latteral movement while allowing vertical movement as well as arresting axle rotation.

Edited July 19, 2023 by IamScotticus

[KnifeySpoony]

I can't help but think that it will twist somewhat with body roll - but maybe the central Aframe bush is soft enough to deflect that much easily. Maybe I should put a gopro underneath my car and find out. Better to check on a car with soft rear spring though.

[CarlB]

The twisting motion from the car rolling is what breaks the Aframe. Caterham has made rear Aframes with spherical bearings and I do not understand why they changed. 

[MV8]

It isn't clear to me what you are trying to convey Carl. Things twist after they bind.

Edited July 20, 2023 by MV8

[CarlB]

The A frame twist when the car rolls. When the bushings at both ends require more force than the force required to bend the tube, the tube bends. Eventually the force is enough to break the weld or the tube.  

[MV8]

The axle rolls about the A frame center bush. The bush does not allow much rotation without bind. When it binds, and one outer A frame bush is fully loaded up and the other down, the tubes are then loaded in bending which eventually fatigues the metal. The welded area is thicker and does not fail first. Thinner tubing next to the weld is where the fatigue occurs and it fails after enough cycles but it would be close to the weld. Commonly referred to as "notch sensitivity". The fix is not a stronger A frame.

A spherical center addresses the issue better than having a worn out center bush or eliminating any roll. An A frame center mount spherical instead of a bush allows the roll without bending anything. Depending on how much roll will be allowed, sphericals on the ends of the side links would be better still. The outer A frame bushes do not need to be converted to spherical to eliminate the bind.

The 90s suzuki sidekick also uses side links and an A frame. The A frame is on top and the links are below, but it use a spherical on the A frame and it doesn't break normally, despite a lot more travel and articulation.  

[CarlB]

I think we are saying very close to the same thing about the problem. You gave a more detailed description. You could fix the problem with rod-ends or spherical bearings at the chassis mounts, but you would have to change the mounting bolts from horizontal to vertical. To put the spherical bearing at the outer end can probably be done with a custom A-frame. The GM, I think, G body (1980s midsize sedans) used the triangulated four link as well. Croc has a Holden race car with the same suspension.   

[EdWills]

It was very interesting reading this link regarding 'A' frames, and also the technical notes from David Kaplan (of DSK).  It has been well documented that, quote:  'Colin Chapman realized that the price of the Series 1 chassis frame was becoming too expensive to buy in.  If the design could be simplified, the cost would be reduced. The original drawing by Ian Jones for the Series 2 is first dated November 16, 1959 but the model wasn't listed until June 1960.  The 'A'-frame was not a Chapman invention, but on the Seven it was employed in an economical and effective way'. (paraphrased from Jeremy Coulter - 'Lotus Seven - Collector's Guide').  In Len Terry's book, he advises that the 'A' bracket layout that he incorporated on his Terrier Mk.1 'was inspired by that of the G Type ERA though the latter in fact had de-Dion suspension'.  The ERA had the 'A' frame on top of the De-Dion axle, and the trailing arms were on the bottom.  Terry overcame this by inverting the linkage to lower the roll centre on his live axle car..  "The A-bracket was ball-jointed to the underside of the final-drive housing.  This variation has since been copied for several other cars, perhaps the best known being the original Lotus Cortina and later models of the Lotus Seven'.  Len Terry worked as a draughtsman for Lotus during 1958 and 1959, then again at the end of 1962 to 1965 as chief designer. Note:  The Lotus Cortina location for the apex of the A-frame had a swivel with a grease nipple, and used a spacer tube and rubber bushes inside the swivel where it attached at the axle bracket.

 

When I had to purchase a new chassis frame from Arch/Caterham to replace my badly damaged unit (mid to late 1970s), it came with a new 'A' frame.  The original had been braze welded, but Arch completely fusion welded my new 'A' frame. They had also machined the slots to take circlips to secure a spherical bearing if required.  I've mentioned in another post how Caterham tried a couple of ways to secure the apex of the frame to the axle (ball joint, sliding ball joint), but a letter from their technical director advised that perhaps Chapman had it right, as - especially for a road car - the rubber bushes provided the necessary flexibility, providing the best suspension with this layout.  This assumed that the rubber bushes would not be contaminated with oil or grease, and were kept at the correct tightness. I am not sure if the Ford axle suffered as badly as the Standard Triumph unit, thus the axle strengthening plate may not be as necessary?  Comments please from those with a Series 3, without the steel strengthening plate or other mods added to prevent the Escort axle from twisting. 

 

One poster inquired why the radius arms on the Series 2 (and my Lotus factory built Seven Series 3, shipped to Canada) had the bend in them.  According to Tony Weale, 'the Lotus built Series 2 Seven radius arms are cranked at their trailing end to provide clearance for the handbrake linkage of the Standard axle, and these parts continued for the Lotus Seven Series 3 though the handbrake linkage had moved'. No doubt Lotus had many spares of these, so used them up on the Escort axle cars.  Caterham manufactured straight radius arm as replacements.  Lotus also used cranked radius arms on the Series IV where they connected to the rear of the chassis frame.

 

As you may note, all of my information comes from the many excellent books on the Lotus Seven by various knowledgeable writers.  I purchased these books when they first came out as my car had been modified somewhat by the second owner (I am the third), and I wanted to know what was correct and what was not.  These same books are now ridiculously expensive as it seems that they have not been republished.  I really appreciate the input of the many knowledgeable contributors to this forum as well as the extensive library of lotus@se7ens.net originally started by John Watson.  John D. and I have noted discrepancies in some of the books, but former Lotus factory personnel have put things straight on some of the queries we had.  Bill F

 

 

Edited December 16, 2023 by EdWills
additional information

[EdWills]

Part 2.  My apologies if I have gone on too long on this subject, but I forgot about an article I found in one of the Lotus Seven Portfolio books I have.  In a magazine article from a U.K. periodical included in the Portfolio book, probably around 1969 or so, the author was advising how to recognize a 'genuine' Seven and what to look out for in the way of repairs and possible frame structure problems.  He noted that some* owners of Series 2 and 3 cars made a fix at the location where the 'A' bracket attaches to the chassis frame just to the rear of the seats.  They fitted a 1 inch x 18 gauge round tube with threaded inserts (1/2 inch - 20 UNF) in each end, and bolted this between the 'A' bracket ends with the rubber Metalastik bushes fitted to the ends of the bracket. Longer 1/2-20 bolts grade 8 or better (AN?) would be required, plus shims or washers to centre the 'A' bracket in the frame.  (I purchased a packet of various thickness shims some time ago from Dave Bean with 1 inch o.d. and 1/2 inch i.d. - very useful).  The 18 gauge hollow tube secures the ends of the A bracket to the rear of the chassis, and also adds some strength to the area of the frame just behind the seat squabs.  As mentioned above, in order to save costs on the Series 2 (and 3 with an almost identical chassis frame), the tube fitted behind the seats at the bottom of the chassis on the Series 1, was removed on the Series 2/3.  By adding the round tube between the front of the A bracket locations, it adds some strength to this area almost triangulating the tubing behind the seat back again.  For authenticity, if it is required to return the car back to factory trim, the tube can simply be unbolted as it is not a permanent structure.   *Note: I have never seen a Series 2 or 3 Seven fitted with this modification or viewed any photos of it, but in the Lotus Seven book by Dennis Ortenburger, he shows a Series IV with this modification added to the rear suspension.  I do not know how many 'some' are, but the author of the article seemed to have knowledge of it!  Bill F.

Edited December 19, 2023 by EdWills
info added

[MV8]

That would make a good chassis brace, adding no un-sprung weight. It could also be threaded on one end to eliminate the need for shims. A telescopic fwd mount torque arm with standard trailing arms would work well and fit but obviously not for an original seven.

[EdWills]

Apologies, drawing would not load 

Edited December 19, 2023 by EdWills

[MV8]

Thanks for the drawing. No need to change that to prevent bind on the S1 if the swivel is kept greased. Do you have a good drawing or photo showing the swivel/ball assembly on an S2? This is the best photo I have.

[EdWills]

Hi MV8.  These are the only diagrams that I could find.  My thanks to David Kaplan for his series of bulletins, now posted on SimpleSevens.org (John D. thank you also), and to John C. and Scott (Iamscotticus) for their advice to post these.  

 

Image 1. This is the lower 'A' frame produced by DSK.  It places the spherical bearing in the correct plane to prevent side-to-side movement of the axle without over stressing the bearing.  When Caterham promoted the spherical bearing as a possible solution to premature failure of the rubber bushes, they placed the bearing in an upright plane (to fit their existing 'A' frame), and used bearing spacers and circlips either side of the bearing.  I am not an engineer*, but I would think that this will place more stress on the bearing causing faster wear on the bearing ball and its housing, and I'm guessing that it was never designed to be used in this way?  Len Terry on his Terrier racing car, located the spherical bearing in the same orientation as DSK.

 

Image 2.  This shows the tubular brace designed to connect the sides at the rear of the chassis, just behind the seats on a Series 2 or 3.  As MV8 advised, it does not add any un-sprung weight, but does provide extra strength to the 'A' frame and the chassis in one go.  It can be removed if required to return the chassis to as built.  Longer bolts and perhaps washers or shims may be required to centre the 'A' frame.  Note: The drawing should read 'mild steel' not 'mile steel'. 

 

Image 3.  This is a diagram of the original Lotus Cortina method of locating the 'A' frame on the car.  The swivel is located with rubber bushes, and provided with a grease nipple to lubricate the swivel.  The diagram is from the Autopress Ltd. Ford Cortina 1967 - 68 Autobook Workshop Manual for the Ford Cortina 1300, 1500 and 1600cc  including Lotus 1967 - 68 by Philip H. Smith.

 

*Note: I am not a draughts-person or good at drawing, so please excuse the add-ons in image #2. Oh well!

 

Cheers,  Bill.

Edited December 25, 2023 by EdWills
spelling error

[EdWills]

One more image to add from Tony Weale's book on the Lotus Seven as requested. by MV8..  It shows the mounting of the spherical bearing mentioned in the post above.

Happy New Year to all.

[EdWills]

Hi MV8 and Scott.

Attached is Arthur Mallock's take on locating a solid rear axle in a sports car using a Watts linkage system and disc brakes. The very early Mallocks were fairly similar in appearance to a Lotus Seven.  Arthur Mallock advised me that he worked on a number of Sevens in the U.K. for customers who raced their cars.  As noted previously, this modification may not/would not be sanctioned by various U.K. motor racing bodies.  Unfortunately, the extreme right hand side of the notes are cut off when being copied. This is due to U.K. paper size being different to that in North America.  Mr. Mallock used long, almost parallel radius arms either side of the chassis to locate the axle fore and aft, but the attachment points are not shown in this diagram.  Cheers B.

 

 

.

Edited January 10 by EdWills

[MV8]

Thanks. I have not modeled the Mumford. I wonder if the roll center stability is both vertical and lateral. The height adjustment is nice to complement an existing IFS RC height. Here is the full page with the missing text. IRS components fitted to a three link version of a De Dion (two lower, one central short upper to a rigid rear tunnel) and located by a Mumford seems to be ideal all around, short of a five link IRS or semi-trailing config for heavier/higher power applications like the rx7.