There have been some unusual engines used in racing and road transport. Setright’s excellent book (1) covers many of these. Many of them had significant problems though, or did not lend themselves to mass production.
One very distinctive and still very relevant race engine that has slipped from many engineers’ memories is the original Polimotor however. It remains worthy of research for two reasons. First, despite its age, the original engine still has many significant technical merits decades after its conception. It is notably light compared with contemporary engines. Second, its creator has persisted with research into novel materials and manufacturing processes, and a new engine is now being developed.
In the mid-1980s this car, equipped with an engine made largely of polymers, caused a real stir and scored top-three finishes against OEM competition (Courtesy of Mark Windecker Photo)
Matti Holtzberg has been involved in this engineering adventure for decades, and he says various polymer companies have come and gone from the project, having used IC engines as a method to prove their materials and manufacturing processes which Holtzberg has helped develop for more than 40 years. The list of components made wholly or in part from polymers in that original early 1980s engine development programme remains impressive even now. There is an interesting SAE Paper (2) covering the engine from a materials point of view.
The Polimotor 1: the early days
The original Polimotor engine, the Polimotor 1, was a 2.0 litre, four-cylinder, naturally aspirated race engine producing around 320 bhp at 9500 rpm. It weighed 76 kg, whereas the all-metal version of the same engine weighed 143 kg. Such was the list of polymeric components used in the engine that the SAE Paper gave only the shorter list of those that were still in their original metal.
The crankcase and cam covers were made from graphite-reinforced polymers, but it is perhaps the list of moving components that shows the adventurous nature of the material developments. In the valvetrain, the spring retainers, intake valve stems and the inverted bucket-type flat-faced followers were made from polymers.
In common with the later aluminium followers used in some production engines, the Polimotor followers used a metal shim on top of the follower that doubled as a wear-resistant surface and a method to adjust valve clearances. Some companies were already experimenting with aluminium spring retainers and followers, and Holtzberg says that, during development of Polimotor 1, “Retainers and followers were the easiest parts to do and were a good substitution for aluminum”.
At the same time as the four-cylinder engine was being developed, some components were tested on pushrod engines and polymer rocker arms and composite pushrods were among those successfully tested.
For such an experimental and adventurous engine, it had a very successful racing career. Installed in a Lola T616, it took third-place finishes at Lime Rock and Elkhart Lake in the 1984 IMSA GT Championship, beating competition from Mazda, Porsche and Buick. A decade later, the Polimotor 1 was competing again in the British Hillclimb Championship, scoring four victories.
The Polimotor 1 also made extensive use of polymers in the cranktrain. The piston would clearly not be capable of withstanding combustion temperatures but, such was the spirit of experimentation, a set of pistons was made from polymer and, according to Holtzberg, “The full plastic piston ran for 8 or 9 minutes on the dyno before a hole was blown in it. When we took it out, it looked perfect from the bottom ring downwards.”
A composite cylinder head with an aluminium cam and tappet carrier and deck plate was typical of the adventurous Polimotor 1 engine (Courtesy of Cascone Auto Sport)
So, a piston featuring a polymer skirt was the next development, leading to a significant reduction in mass compared with an aluminium part. The weak point in the design was the bond between the plastic skirt and the aluminium crown.
Piston pins produced with a carbon fibre-reinforced polymer (CFRP) core and a thin metal skin were successfully tested in the Polimotor 1, and such components could be made cheaply as an injection moulding, but in many forms of racing such components are specifically banned. The eminently sensible idea of a multi-material, stiff, lightweight piston pin falls foul of rules mandating that pins are made from a single piece of steel, but which place no limits on the complexity of such a steel part.
Polymer piston rings were also used in the Polimotor 1; this was a technology that engineers had been trying to develop for many years, both for racing and production. In the Polimotor engine, the top ring remained metal, but the second and oil-control rings were polymer. The retention of a metallic top ring is important for heat rejection, as the heat path through the polymer skirt has a much higher thermal resistance.
Non-metallic con rods, according to Holtzberg in a masterpiece of understatement, “were always a challenge”, but were successfully made through a process that he referred to as triaxial braiding using continuous filament. I know of only one other attempt to produce a serious racing con rod from composites, and it was not a success. Holtzberg says, “Polymer composite rods worked well, but especially if the piston and pin were light”. In this context he explained that the light pistons and pins were those manufactured by his methods.
It is interesting to note that the only serious mechanical engine failure during the racing history of the Polimotor 1 was a steel con rod from a well-known manufacturer, although Holtzberg refers to this as “probably an assembly problem”.
There is a train of thought in design that you should be adventurous in whatever you are trying to achieve and fix the faults – there probably won’t be as many as people expect. That was certainly the case for the Polimotor 1: the list of polymeric components was not limited to those engineers might have thought were guaranteed to be successful.
The original Polimotor engine relied heavily on Amoco Chemical Company’s poly-amide-imide (PAI) thermoplastic materials, which are a family of polymers with high-temperature capabilities. Polymers generally, compared with metals, have far lower properties in most regards, but the Torlon PAI as used for the Polimotor 1 is closer to the properties of metallic materials in many regards than almost all other polymers available at the time the engine was conceived, and it is still the case now.
Matti Holtzberg with the moulded composite cylinder block from the Polimotor 2. This manufacturing technique produced parts quickly and with inexpensive tooling (Courtesy of Polimotor)
PAI is an amorphous polymer with good resistance to wear and creep. It has good chemical resistance and can be used up to 260 C. It has a low coefficient of expansion for a polymer. Of more relevance in the modern world, it also has good dielectric strength for electrical insulation. I first used PAI more than 20 years ago for a hybrid system application, precisely because of its combination of properties.
Bolted joints
Holtzberg says he simply “looked differently at why metals worked” for engine components and thought that, for many of them, polymers could be made to work as well. Unencumbered by the same experience and traditions as engine designers used to dealing solely with metals – steel, cast iron and aluminium making up almost all the mass of every engine – Holtzberg looked at the loads involved and the points where a polymer component might struggle to handle those loads.
In the 1980s, with stresses in engine components being lower than now, it was easier to make material substitutions – as Holtzberg puts it, “In those days, plastics could work for many components.” With a practical background, he realised that putting durable, heavily cyclically-loaded threads into polymer parts was going to be problematic, especially for main caps and cylinder head joints. He says, “Threads imposed a lot of load on moulded materials” and therefore, where possible, they were eliminated, most notably by using the technique of through-bolting to produce the main load-bearing structural assembly of the engine.
This effectively ‘sandwiches’ the whole engine, from the cylinder head to the crankshaft main bearing caps, and was used so that the block was in compression and none of the plastic components had threads subject to combustion loads. That was around the time Rover was doing the same on its K-series engine, and Holtzberg’s use of the technique was a simple and pragmatic solution to a foreseeable problem. The use of through-bolting is applicable only to inline engines, and is still used in some highly stressed all-metal race engines.
At other points on the cylinder block where a thread was unavoidable, Holtzberg used a threaded sleeve with a longer thread engagement than a normal thread straight into a typical metallic casting. As an example, he said that with a typical M6 thread he would create an insert with an M8 or M10 thread on the outside and of a length to accommodate 2D-4D of thread engagement internally. The larger diameter and coarser pitch of the external thread gave resistance to initial thread stripping, and the long length allowed the load to be spread over many more thread pitches than would be typical in a metallic part.
The latest project is Polimotor 3D, which explores metallic AM parts. These 3D cams have intricate weight reduction details and will run DLC-coated (Courtesy of Polimotor)
Where neither of the above methods were possible, he said the threads were drilled, tapped and fitted with coarse-thread helicoil thread inserts.
According to Holtzberg, at the time of Polimotor 1, Torlon had “[mechanical] properties about half of those of 356-T6”. A356-T6 was the predominant high-strength aluminium casting alloy used in motorsport at that time. He also noted that batch-to-batch variability of Torlon materials was significant at the time. “If you were machining parts and the swarf was like powder, the material had been over-cured and the parts would not be strong enough,” he says. “Usually the swarf came off nice and the parts would be good.”
End view of the 1980s Polimotor 1 engine, which shows how little of the external structural assembly is metal (Courtesy of Polimotor)
More than 35 years after the original Polimotor 1 engines were produced, Holtzberg has recently been approached to produce a small batch of new ones for use in historic sportscar racing. Amazingly, the tooling to produce some of the major components of Polimotor 1 still exists and is usable.
The Polimotor 2
The second iteration of Polimotor engine is based on a greater use of composites and focused on a smaller set of components with which to prove the advantages of composites. This time, in partnership with Solvay, a range of processes from ‘composite casting’ of CFRP bulk moulding compounds to some relatively early work on carbon-reinforced additive manufacturing (AM). It proved to be a bridge between more established processes and the emerging world of AM that has really come to the fore in recent years.
The Polimotor 2 was never used as a race engine, although it accumulated around 60 hours of dyno testing. It was however a success in terms of proving novel manufacturing techniques.
The initial idea was for a cylinder block to be moulded from fibre-reinforced composites, which has been referred to as a composite casting technique. While such materials have gained very wide acceptance in automotive manufacture for inlet manifolds, covers and so on, the idea of a large-scale, structural part had not been seriously explored. The accepted methods for block manufacture at that time were restricted to casting, either sand-casting in low volumes or die-casting in higher volumes. For both methods, the tooling incurs significant costs.
For injection moulding of composite blocks, the tooling was relatively cheap, and the speed of production could have been competitive with prototype production – demonstrations have shown it can produce an engine block every 10 minutes from a single mould. For racing purposes with larger budgets, there is also the possibility of laying unidirectional composite cloth into the mould to reinforce parts in critical areas to increase strength and stiffness. This can be achieved by addition bonded CFRP patches applied to metal castings, but that is not as cost-effective as doing the same as part of the manufacturing process with a composite part.
The Polimotor 3D and the future
While Holtzberg is not the first to realise the benefits of 3D printing for small-volume manufacture, he is frustrated that people aren’t yet doing more with it. He certainly hasn’t given up with polymer materials, and the Polimotor 3D is Holtzberg’s latest engine project; it remains a polymer-rich engine but also looks at non-polymer components.
However, Holtzberg is looking to revolutionise the manufacture of metallic components too. There are several avenues to explore here, such as simple reverse engineering of obsolete components, and redesigning components from first principles to produce something fundamentally better by eliminating the constraints that come with traditional manufacturing methods.
For a 3D-printed engine block produced by an enhanced extrusion process known as pellet extrusion, Holtzberg has shown that a four-cylinder engine block can be produced in around 42 hours. Compared with casting, that is slow, but in the 26 weeks that is often a typical gap between placing an order for a casting and receiving the first parts, Holtzberg’s printed block would be ready for machining two days after receipt of an order.
That introduces a step change in the rate of development, and the process benefits from having zero tooling charges and the ability to make changes extremely quickly. Several design iterations can be produced and machined, built into engines, tested and assessed using this method before the first conventional castings are available.
Many engineers will ask, “What about production?” The injection-moulded composite block is the obvious production method to use for higher quantities and faster rates of production, at least while the present rates of production by AM in useful structural materials remain as they are.
Holtzberg says the concept of pellet extrusion is a development to improve on fused deposition methods, which use filament extrusion. Where filament extrusion melts a specially produced filament of material, specifically aimed at a certain type of AM – fused deposition, also known as fused filament fabrication – pellet extrusion takes advantage of the ubiquitous raw materials used by injection moulding companies worldwide. Such pellet materials are produced in much greater quantities than filament, so are far less expensive and are available in a very much wider range of plastic ‘alloys’.
Besides cylinder blocks, many other 3D-printed parts are planned for the Polimotor 3D, which is planned to appear next year in sportscar competition.
Metallic components
One of the first additively manufactured metallic components that Holtzberg has looked at are camshafts. The aim is to show that near-net-shape blanks can be produced and used with minimal machining and, where desired, coating too.
Many of us understand the considerable benefits that 3D-printed parts could bring to producers of premium powertrain components. In the case of the camshaft, there are certain fixed areas of geometry (bearings and lobes, for example) that we need to retain. However, where camshafts see ‘double duty’ as integral parts of the lubrication system by providing oil to the cam bearings and the contact between lobe and follower, there is scope to reduce weight, part count and reliance on seals.
It is a common concept to use the camshaft bore, in conjunction with a tube carrying multiple O-ring seals, as an annular conduit for lubricant. A camshaft blank produced as a single part with integral thin-walled passages would represent a real improvement. Holtzberg has produced parts that have been tested in uncoated format and further parts have been successfully DLC-coated. The scope to produce fundamentally simple parts that would be difficult or impossible to produce by conventional production methods is what the Polimotor 3D aims to show.
Of course, the benefits of using lightweight AM parts extends into the world of electrification as well.
Summary
Holtzberg’s work is historically significant in the field of engines. The Polimotor 1 was many years ahead of its time. A number of senior race engine designers from well-known companies followed the path to Holtzberg’s workshop to see and hear the engine running in anger. Among these were well-respected figures such as BMW’s Paul Rosche.
We like to think that motorsport represents a pioneering arena and a simmering cauldron of clever ideas. However, the reality that also applies to a greater extent in mainstream automotive industry is that we are inherently conservative, now more so than ever. To take a risk on a new technology is hard to convince people of its benefits. We do it when we are forced to by regulation, but we will rarely do it unless the benefits are substantial and the risks are low.
The Polimotor 1 and the variants that followed showed that there was, and remains, the possibility to produce a reliable and competitive engine from unusual materials and which is much lighter than the conventional metal alternative. It is true to say that today’s much more highly stressed engines offer less scope for such adventurous material substitutions, but that scope still exists where regulation allows.
References
- Setright, L. J. K., “Some Unusual Engines”, Institution of Mechanical Engineers, 1975, ISBN 0-8529-8208-9
- Gaudette, E. P., “Plastics Within the Internal Combustion Engine”, SAE Paper 850815, 1985
DATASHEET
POLIMOTOR 1
Inline 4
95.2 x 71.12 mm = 2025 cc (123.7 cu in)
Naturally aspirated
Gasoline
Carbon fibre thermoplastic block and crankcase
Cast iron liners
Five plain main bearings
Steel crankshaft, four pins
Steel con rods
Aluminium/polymer pistons, three rings
Belt-driven DOHC
Four valves/cylinder, one plug
38º valve angles
36.8mm intake valves, 31.8mm exhaust
Butterfly throttle
13.5:1 compression ratio
Maximum rpm, 10,000
SOME KEY SUPPLIERS
Block and head machining: in-house
Liners/liner coating: Advanced Sleeve
Crankshaft: Moldex
Con rods: Carillo (for steel parts)
Con rod bolts: ARP
Pistons: Mahle
Piston rings: in-house
Piston pins: Mahle
Main bearings: Clevite
Big-end bearings: Clevite
Thrust bearings: Clevite
Circlips: Mahle
Valves: Ferrea
Valve seats: in-house
Valve guides: in-house
Valve springs: PPE
Spring retainers: Ferrea
Camshaft drive components: Gates
Camshafts: Crane
Cam followers: Del West
Cylinder head seal: in-house/Cometic
Fuel injectors: Injector Dynamics
Ignition: Ole Buhl Racing
Engine management: Ole Buhl Racing
Fuel pumps: Bosch
Data acquisition: in-house
Throttle bodies: in-house
Water pumps: in-house
Oil pumps: in-house
Oil filters: in-house
Air filters: in-house
Exhaust: in-house
Wiring loom: in-house
Fuel: VP