"Monocoque"

[ardee_selby]

Gary,

Interesting take on things.

Been musing a bit about Lotus and their contemporaries...

When parallels between Elites & Rochdale Olympics were discussed, Dave-M posted the following, which I think stands a repeat:

"I have two Rochdales undergoing restoration at present, a GT and an Olympic Phase2. The Phase2 I have owned since 1992 and was used regularly until restoration time came along.
It's a true fibreglass monocoque using a suframe made of 1" tube to mount the front suspension and the underside of the shell is very similar to the Elite.
The GT and the Olympic were designed in the main by Richard Parker who was then poached by Colin Chapman to work at Lotus, so the connection between Lotus and Rochdale is a little more than just fibreglass. If any of you are interested in the Rochdale here is the club website :- http://www.rochdale-owners-club.co.uk/"

This webpage (http://www.rochdale-owners-club.co.uk/olympic.htm) is an interesting read, particularly in view of Richard Parkers' subsequent involvement with Lotus.

e.g. ""the rear suspension and the gearbox mountings bolt directly into the glassfibre body without any steel reinforcements"

Why was he "poached"? Was it due to his work as a composite pioneer? (I have read (somewhere) he actually worked on suspension designs on Types 20 & 30)

In any event..there are those that are still flying the monocoque flag!

http://www.vord.net/cars/rochdale/rochdale-2004.html

Cheers - Richard (You there, Dave?)

[garyeanderson]

Some more lame thoughts of mine.

RD

all good stuff, but the monocoque Elan is a roadster with special problems. Was the Elan body strong enough? My guess it was pretty close as engineered. The real reason (in my gracious opinion) that things developed as they did was cost and ease of production. if you think about it, if you took the front and back halves of the Spyder chassis and left out the middle backbone then you would have both of your subframes. My guess is that Ron drew it out in a similar fashion and costed the project and couldn't get the numbers to work. The folded sheet steel was just too easy and cheap to ignore if the project was going to move forward any time soon.

Gary

[Elanintheforest]

Totally agree with the cost and ease of production conclusions. You can buy a new Elan shell today for around ?2500 and a chassis for around ?1800. An Elite shell (or Chassis Body Unit (CBU) as they are known) is also available for ?14000. The Elite is double skinned all over the place, and the 'filling in' of all the normal space underneath makes them very difficult to maintain. From underneath, you can change the engine and gearbox oil. Everything else on the engine has to be done from the top.

The differential is bolted straight onto fibreglass in a small open-at-the-bottom box structure. Not too bad to get in and out, much like the Elan, but you often have to do that to get the brake pads out, adjust the brakes or renew the handbrake cable if things are a little seized.

There are many places where access is incredibly tight, such as the holes to access the nuts that secure the bumpers to the body. I simply can't get my hands in there, by a long way! Lotus must have employed a few dwarf gynaecologists to put these cars together.

Accident repairs must also be considerably more expensive with the Elite. I've wrecked a brand new chassis on an Elan by clipping a kerb. That would cost less than ?2000 in todays money to fix, plus ?1500 or so in labour if you had to pay someone to do it. If you did the same in an Elite, the front of the car would have to be taken apart, and probably a new subframe bonded it. The last job is pretty specialist, and is around ?6000 to get sorted, plus the taking apart / putting back together of the car, so maybe ?7500 for the same job.

There are also the issues of excessive harshness, noise and vibration, all of which are features of the Elite...the Elan is a luxury car by comparison. I'm not sure why, or if it has anything to do with the monocoque vs chassis approach, as both cars have their suspension, steering, engine and gearbox isolated from the body or chassis with rubber. Nobody has solved those issues with the Elite in over 50 years, whilst the Elan has never had problems in those areas.

I would have thought that, with the advent of Kevlar technology, a 'plastic' monocoque car would be a viable option again, at least in the exotic machines, if not in the mass produced market. But for now, the Elite and the Olympic remain as the first and last experiments in this area, and are both all the more fascinating for that!

Mark

[ardee_selby]

<snip> There are many places where access is incredibly tight, such as the holes to access the nuts that secure the bumpers to the body. I simply can't get my hands in there, by a long way! Lotus must have employed a few dwarf gynaecologists to put these cars together. Mark

Maybe "DiG" had owned one?

Cheers - Richard

[cabc26b]

Ardee wrote

Been musing a bit about Lotus and their contemporaries...

Maybe look a little closer to home is what Frank Costin was doing with monocoque designs for Marcos and Costin-Nathans . Not a hard leap to make the elan bodyshell work.

George

[pharriso]

From Colin Chapman?s Lotus (1989; Robin Read, inputs by Mike Costin):

"Chapman was doing double duty trying to keep Lotus Cars and Team Lotus above water. While he busied himself arranging for his racing engines to swap ends of the car, and at the same time planning for the production of his first sports coupe, he was forced to delegate design of the Type 17, the last front-engined Lotus racer. He specified MacPherson struts both front and back of the 17. The front design as implemented was a disaster on the track and had to be scrapped. The fallout was that Chapman spent much of 1959 totally absorbed in salvaging the Lotus reputation from the Type 17?s problems. This in turn necessitated that he delegate initial design work on the next sports car on the drawing board entirely to Ron Hickman. It was to become the Type 26 Elan.

Initially, Hickman set out create the Lotus 26 as a replacement for the Lotus Seven, which had been an impractical (sunny days only) specialty car with minimal benefit to Lotus? bottom line. All initial design for the Type 26 was based on a monocoque body shell with two doors and open top. Nobody had ever built one. In spite of the Elite problems, monocoque was still the preferred Lotus structural design, and perhaps lessons learned on the Elite would help make a better car. But as the Elite disaster continued to unfold, by 1960 it was necessary for the Type 26 to become the replacement for the Elite. Only a big success for the Lotus 26 could save the company.

By mid-1960, Chapman began to make noises about a chassis for the Type 26 to reduce risk in this must-succeed venture. He also became enamored by the independent rear suspension of the Triumph Spitfire, although he wanted a better version for the Type 26. But it was still too early for final admission of defeat on the original design; Ron Hickman continued to try to prove a design for a monocoque plastic structure for an open Type 26. As it got closer to the final scheduled testing dates with no breakthroughs in sight, the engine and suspension engineers went begging for a test platform. Chapman concurred and designed the backbone steel frame in a weekend. It subsequently was used for 50K miles of testing under a fake body shell.

By November 1960, Hickman threw in the towel and suggested they keep the test frame as the basis for the Type 26 chassis. A volume cost estimate of 10? each for the 75 lb. frame sold the idea. Any alternative would be much more costly. Hickman continued his overall design efforts and got approval from Chapman for the pop-up headlamps, a functional gimmick that would distinguish the Type 26 from its competitors. Chapman thought the Type 26 should be called the Elite again, but Hickman preempted him by having a naming contest in which the only acceptable name was Elan."

[garyeanderson]

The replacement subframe

[garyeanderson]

So you thought this topic was dead? I have a few question to ponder. I have been thinking a bit about this since John started the topic.After looking at a few Elans and subframes I really wonder what it offers in terms of percentage. How strong in terms of twist is a box open at both ends? I don't know and I am sure it does what is needed but not much more. Its not a true box but more of a quadrilateral trapezoid and that may help some but it still leaves me wondering. Lest you think I am picking on the stock subframe, what does the "26r" type of mods offer. Spyder claimes 50% increase in stiffness, 50% of What? One of the books offers 4500 lbs/degree of twist for the Elan (race car design?)

http://books.google.com/books?id=_p-6jr ... bs&f=false

for the chassis, it also offers a bit on the question of the open box backbone subframe.

http://books.google.com/books?id=_p-6jr ... an&f=false

So if it's really 4500 lbs/degree for just the 75 lbs of 18 gauge steel and a few spot and mig welds, what is the total with the type 26 body attached to it with the 16 bolts?

Gary

[GrUmPyBoDgEr]

So you thought this topic was dead? I have a few question to ponder. I have been thinking a bit about this since John started the topic.After looking at a few Elans and subframes I really wonder what it offers in terms of percentage. How strong in terms of twist is a box open at both ends? I don't know and I am sure it does what is needed but not much more. Its not a true box but more of a quadrilateral trapezoid and that may help some but it still leaves me wondering. Lest you think I am picking on the stock subframe, what does the "26r" type of mods offer. Spyder claimes 50% increase in stiffness, 50% of What? One of the books offers 4500 lbs/degree of twist for the Elan (race car design?)

http://books.google.com/books?id=_p-6jr ... bs&f=false

for the chassis, it also offers a bit on the question of the open box backbone subframe.

http://books.google.com/books?id=_p-6jr ... an&f=false

So if it's really 4500 lbs/degree for just the 75 lbs of 18 gauge steel and a few spot and mig welds, what is the total with the type 26 body attached to it with the 16 bolts?

Gary

As far as I know the box section is inherently a very stiff structure torsionally; more so than let's say a round tube but less so than a triangle.
Unfortunately once that structure begins to buckle practically all of its stiffness disappears.
A cardboard model would provide a simple example to play with.
It is therefore important that the sides of the box section are kept as flat as possible; kinks or bows will increase the tendency to twist & buckle.

The Spyder backbone would be a relatively weak structure torsionally if it were made merely of square tubing but the flat steel sheet which is point welded to those tubes make it the stiff structure that it is.
Its maximum possible stiffness would be greatly improved if the weld was continuous.
If those flat sheets were formed in "waves" perpendicular to the square tubes, as in corrugated cardboard they would make the structure even stiffer provided they were welded along their whole length (a lot of welding).

John

[elansprint71]

As Gary's photo shows, the central section is not simply an open-ended box, there is a flange with radiused corners which will provide an enormous amount of resistance to twist. As he correctly postulates an open-end box (particularly with such thin walls relative to the box dimensions) will have little resistance to torsional distortion, unless, as is the case here, it is supported by other components.

Ageing Engineer shambles off in search of slide rule...

[garyeanderson]

I got a better photo of the front end of the box section that I am questioning, The back bulkhead looks to be an effective use of that steel. The front seems to be compromised by having the bulkhead disappear in the middle (a doubler would be a better term there). This compromise is dictated by the need to get the tail shaft of the box stuffed in there and to allow for some movement of the back end of the box. Spyders version is also the same in that the front bulkhead is minimal. My guess is as Pete said, the forces are taken else where and the "else where" is the fiberglass body.

Gary

[CBUEB1771]

As far as I know the box section is inherently a very stiff structure torsionally; more so than let's say a round tube but less so than a triangle.

I'll have to dig out my stress and strain textbooks to quantify this, my copy of Roark is at home and not in the office. It is hard to imaging a box section being more torsionally rigid that a round tube. I have not seen many drive shafts with a square section. I am happy to be proven wrong.

[RogerFrench]

In Robin Read's excellent book "Colin Chapman's Lotus" the story of th Elan's difficult gestation is very well documented. All ideas of load-bearing plastic was abandoned well before production started.
Best Lotus book you've never read.

Does it also explain why Lotus chose to bond body & chassis on S1 Europas'? And was it being leant on by insurance companies that brought about the change?

Cheers - Richard (BTW re: "I'm going to take myself off to an island and not going to post for a week". Your week isn't up yet, so s*d off! )

Sorry, I have only just read this question.
Robin Read left Lotus before the Europa came along, so wouldn't have been in a position to comment.

[GrUmPyBoDgEr]

As far as I know the box section is inherently a very stiff structure torsionally; more so than let's say a round tube but less so than a triangle.

I'll have to dig out my stress and strain textbooks to quantify this, my copy of Roark is at home and not in the office. It is hard to imaging a box section being more torsionally rigid that a round tube. I have not seen many drive shafts with a square section. I am happy to be proven wrong.

Russ, my battered copy of Roark is sitting on the shelf here in my new den, I won't bother with referring to it but continue on gut feeling.
Round bars & especially tubes are extremely good at absorbing torsion & regaining their original shape without deformation.
Box sections just don't like being twisted & will defend their right to maintain their shape until they buckle, resulting in massive torsional failure & permanent deformation.

Totally non-scientific & sheer gut stuff which I am sure the stress-men here will surely give me a thorough hammering about

Cheers
John

[bill308]

Loading on an automobile chassis/body can be quite complicated. There are weights to be supported like the body and chassis, engine, drive train, fuel tank, occupants and sundry other items not only in their static condition but also in their dynamic condition due to breaking, acceleration, and turning. There are suspension loads in the verticle, fore-aft, and side directions. There are also vibration and impact loads and decisions as to factors of safety based upon 1g and multiple g loads.

A monocoque structure can yield outstanding stiffness and strength for its weight but it must be modified to account for occupant space, aerodynamics, getting in and out, mechanical accessibility, and visibility, among other things.

The problem with monocoques is the resolution of point loads and in the case of a roadster, the open area of the cockpit and the necessary door apertures.

On a GRP (Glass Reinforced Plastic) structure, point loads, like suspension pickups, can be dealt with using localized strengthening, like bobbins, a localized increase in wall thickness, and if necessary the intersection of additional GRP panels, ribs, or substructures.

Overall chassis/body stiffness is more difficult. The body/chassis structure must resist bending up and down, side-to-side, and in torsion. Multiple closed, or triangulated sections, are the general strategy, where one makes each section of the structure as rigid as possible and combines them in a rigid way. There is the engine section, the cockpit, and the rear section, usually separated with bulkheads.

The problem with a monocoque chassis in an open roadster is stiffening of the cockpit area and the scuttle bulkhead where the occupant?s legs must go. The scuttle bulkhead is probably best dealt with by the construction of a hoop arrangement where a solid panel is modified with a centralized hole and additional details to make up for the hole's weakening effect. This substructure can be made quite light and stiff in plane. Race cars can use an outer tubular hoop and an inner tubular hoop concentrically joined together with a welded web. The requirement for doors in an open roadster will tax the structural engineer as the necessary cutouts will severely weaken a simpler structure in both bending and torsion. The Type-14 Elite addresses this issue with a structurally integrated top. This is not a perfect solution as it can?t be perfectly triangulated unless somehow tied in to a roll cage of some sort. Still, the large cross section helps a lot. An Elise addresses this issue with large cross section sills on both sides of the cockpit effectively creating a massive two beam structure.

A monocoque structure is stiff due to its geometry. When one deviates from the ideal geometry, band aids must be added. The Lotus 25 Grand Prix car has no doors but it does have an open cockpit, so an important part of the structure is missing. IIRC, this was overcome by two large section tubes on either side of the driver that were also gas tanks.

The most expedient method of achieving stiffness is a large cross section geometry. A hollow tube of large diameter will be much stiffer than a solid rod of the same weight. Similarly, a hollow box section beam will be much stiffer than a solid beam of the same weight. This is true in both bending and torsion. The idea is to put the material weight were it will do the most good, as far away from the neutral axis as possible. In engineering speak, it?s all about the moments of inertia, both in bending and torsion.

Some moments of inertia (stiffness) relationships to ponder:

???????????????...........................Ix (verticle bending)??Io (torsion)
For a solid rectangular beam:?.................bh^3/12???????????bh(b^2+h^2)
For a solid circular beam:???..................?r^4/4???????????? ?r^4/2

Where:
^=raised to the power Example ^2 is squared; ^3 raised to the 3rd power; ^4 raised to the 4th power
b=the base dimension of a rectangular cross section
h=the height dimension of a rectangular cross section
r= the diameter of a circular cross section

For hollow sections the technique is to subtract the hollow void value for a solid from the larger solid section value.

Example for a rectangular section: Ix solid-Ix cavity= Ix hollow tube ; The difference between solid and cavity dimensions is the combined wall thicknesses.

Note the h^3, b^2, r^4 terms. Small increases here yield big stiffness benefits.

Although dated now, the best book I?ve read on the subject is, ?Racing and Sports Car Chassis Design? by Costin and Phipps.

In the case of the Elan, I suspect the biggest problem was trying to obtain adequate chassis/body stiffness in a roadster. Except for the case of the coupe and +2, there wasn?t a good solution for bending and torsional rigidity available in the allowable development time frame. Likely, the solution for an all GRP chassis/body would have been a combination of a large cross section closed center tunnel and large cross section sills. Because of geometry constraints, the closed sections would likely have had thick and heavy walls. Another possible solution would have been to make the doors structurally rigid with structural, locating locks upon closure.

Because of time and likely financial constraints, Hickman and Chapman adopted the backbone chassis as pretty elegant solution to take out suspension point loads and stiffen the cockpit area of the chassis with a minimum weight penalty. It also isolated the occupants from the high frequency road harshness because the body was not rigidly coupled to the backbone chassis except in selected areas like the rear bulkhead to suspension towers, where the verticle wall forms an effective cross brace. I note that Dave Bean's 26R uses a cross chassis rod for much the same effect on the front suspension mounts.

Holed bulkheads in the chassis interior are effectively rigid hoops at the beginning and end of the central tunnel and limit tunnel geometry distortion under torsional and bending loads.

As a young puppy, I was thinking in terms of structural doors and locks when I envisioned building my own special roadster. This was a motivator for my return to college to get my mechanical engineering degree some years after I finished my stint in the army. I still have that dream. Maybe some day.

Chapman and Hickman were pretty crafty, creative folks.

Bill