Lotus 900 series engine history
Further to my earlier posts on the Lotus engines, a good friend (Tony) passes me a few links to forums where a technical paper was reproduced. I am going to reproduce it here as well. Its answered some of the questions on why the prototype Lotus engine has that huge oil pump on the side and why the pistons are cut as they are.
The article was Published by the Institution of Mechanical Engineers in 1973 and the article features a picture of the test cars, the Viva GT, the Type 62 race car, the Vauxhall Victor and the bedford CF2.
Two sites that discuss this article are as follows:
Here is the transcribe of the paper, (so its easier to read)
In the mid 1960‘s the company recognized the need for a 150hp (112 kNm/s) 203Nm torque engine, to give a low bonnet lone, weight and centre of gravity. Such an engine was not commercially available then or in the foreseeable future, so it had to be specially designed, developed and built. The circumstances surrounding the original concept vitally affected the engine design and should be considered before reviewing the engineering decisions.
The company were about to move form a thirty cars per week London area factory to a purpose built plant in Norfolk, far form traditional sources of material and skilled labour, with a capacity of three times their current volume. Their 1500cc twin cam engine, based on their 116E block with Harry Mundy’s designed head was machined and assembled by sub-contractors, some of whom had already indicated that for various reasons, including obsolescence, that they would shortly (in months) cease to contribute.
Exhaust pollution control was a small cloud on a far horizon, 1974 and 1975 Federal emission regulations and vehicle construction requirements for added occupant protection were, mercifully perhaps, unknown to the Lotus planners. Finally a large part of the company’s considerable skills and limited engineering resources were devoted to motor racing and particularly to the first use of the Ford Cosworth DFV engine, and competition at Indianapolis.
At the material time, the American market for the company’s products was less that 10 per cent and was not a major factor in planning.
The concept study, after using consultants, as well as the company’s own resources, came up with a light alloy 2-litre 45’ angle four cylinder, with four valves per cylinder, having two overhead camshafts driven by a toothed rubber belt. This could grow to a 4-litre V8 with Indianapolis potential. Six cylinders in 120’ form were rejected as too wide, in 60’ form as too high. A project team was then assembled and design work began. By the time the cylinder head design was complete, with the basic outlines of the crankcase and crankshaft assembly, it became clear that the cost-time requirements of the engine program would severely tax the company’s resources while they realized the full potential of the factory move. At this time the Vauxhall slant 4 range was announced and it was found that the cylinder head for the type 900, as the project became known, would fit the Vauxhall block with minor modifications. Development could then economically proceed to a stage where enough would be known about the engine to enable a decision to be mad eon its future. Around this time a need arose in the completion side of the company for a 2-litre sports car engine, for races of 800 to 1600km duration. This then led to the cylinder head development splitting into two forms, a large port version, with an included inlet port angle to the inlet valve of 41’ for racing and a smaller port, with greater included angle of 51’. To give a lower carburetor mounting and this bonnet line. The engine family arising then became:
Type 904 : Iron Block 2-litre fuel injection racing engine.
Type 905 : Iron block 2-litre touring engine.
Type 906 : Sand cast aluminum block 2-litre fuel injection racing engine.
Type 907 : Dia cast aluminum block 2-litre touring engine.
Type 908 : Aluminum 4-litre racing engine.
Type 909 : Aluminum 4-litre touring engine.
3. The ‘Iron’ engines
The type 904 racing program, during 1969 was valuable in so far as it showed no basic weakness in the connection rod, bearing and piston layout, as well as the head.
Problems encountered were timing belt jumping the inlet camshaft, due to insufficient warp (less than 70’) dictated by the position of the accessory layshaft on the iron block; the racing clutch gave some trouble and there was a vibration problem with the horizontally mounted distributor; also the iron cylinder block cracked around the main bearing bolt bosses. A special batch of thicker castings was produced and subsequently applied to all production. Power outputs, shown in FIG 357.1 were found to be competitive and could be maintained reliably for 1600 km races, without major intervention form the engineering team fully, occupied with the 907 engine.
The type 905 touring engine initially gave trouble on the test bed with extreme wear on the side of the connection rod against the crank webs, any bearing distress was an after effect and the problem was not present on the racing engine. It was finally eliminated by fine boring the little end bush instead of honing. Honing too fine did not hold oil. The first engine gave 147hp (110kNm/s) on its first power test with a very inefficient exhaust manifold, dictated by the layout of a production saloon, into which the engine was fitted for road testing. The only problems encountered were failures of the vertically mounted distributor, due to vibration, and staring problems, due to the battery capacity not being adequate for the large engine.
Limited but sufficient emission testing indicated that up to 1973 standards, perhaps 1974, could be met by the basic engine without additional equipment. It was also found that the engine was excessively nosy mechanically despite the rubber timing belt, and the air intake roar was also excessive.
Development of the range of vehicles for which the engine was destined, together with new construction regulations, dictated more power which was primarily obtained by a more efficient exhaust manifold obtained by slight changes to the vehicle layout when 154 hp (115 kNm/s) was reached with a fatter power curve.
By the time (early 1970), the engine showed sufficient promise to justify a decision to proceed to the next two parallel stages:
(a) To acquire a machining and assembly facility to produce the engine from die-castings.
(b) To develop the type 907 die cast engine to a standard suitable for volume production.
Circumstances has changed form those outlined in the introduction, the company now had s flourishing US market, foreshadowed Federal Regulations were causing redesign and delay of the vehicles for which the engine was intended and there was a minor recession in the industry.
4. Development of the basic type 907 engine.
The machining facility selected using CNC machining centers where the castings were rotated around the crankshaft axis under twin vertical spindles necessitated some detail redesign to ensure all faces were parallel to the crank axis and all drilled holes at right angles. A number of holding lugs and jog points were also added, but the redesign amounted to a week’s drawing office time. At the same time, casting drawings were modified in conjunction with the selected foundries. A long-term plan was formulated whereby in-house machining of castings would be followed by setting up a machining facility for major ferrous components, these being crankshafts, connecting rods, layshafts and camshafts in that order.
The type and extent of machining facilities planned for production of the aluminum version of the engine had a considerable bearing on the final design. As schemed originally the cylinder block was a full-skirted design but this was modified to incorporate a joint on the crankshaft centre line for two reasons:
(a) It contributed considerably to solving Foundry problems with the die cast components.
(b) There was a dislike for the need to machine deep main bearing saddles and tenons which would have demanded the use of expensive broaching equipment.
By adopting the integrated bearing cap and skirt castings as a separate unit overcame these problems and undoubtedly restore stiffness to the cylinder block resulting from the deletion of the skirt on that component. Similarly, the removal of the layshaft housing form the main cylinder block casting contributed to easier die casting technique; also as simpler machining layout in the separate component form.
A further example of the consideration given to detail design in the original conception can be seen in the evolution of the camshaft housings. These being separate from the cylinder head permitted the use of smaller and more easily machined castings. Similarly the decision to accept feeding the camshaft in from one end during assembly contributed to cost reduction to the elimination of separate bearing caps and it is believed a more rigid assembly.
A very elaborate machining and engine development program to produce a satisfactory combustion chamber that could be produced in two passes of a single cutter had just been completed when the foundry announced they could produce the original combustion chamber to the required limits, without any machining. The engine development was split into two parts, that for a European specifications engine and for a Federal emission engine. The former was to have fixed choke carburetors, the latter a pair of constant vacuum carburetors feeding via a water jacketed manifold. The hot water emerging form the head to be passed around the manifold and overhead balance pipe en route to the thermostat. A batch of twelve sand cast pilot engines was commissioned and apart form being more noisy mechanically that the iron engines, these units performed with reliability achieving their specific output without and further modification.
Initial road testing of these engines was carried out with other manufacturers’ production touring cars and a 1500kg delivery van operated by the company’s supply division. This van was one of a fleet used for collection and delivery of material throughout the country and received the same running maintenance as the rest of the van fleet. Apart from an insatiable appetite for transmissions, rear tyres and brake linings, the vehicle was trouble free and covered over 222000km
Figure 357.2 shows the vehicles used in the vital road test program. The failure rate of the cylinder head joint was high. The cylinder centers, derived from the iron block limited the seating area between the liner and the block, roughly half the liner was unsupported, i.e. the stroke or 69.8mm. The final solution was to mount the liner in metal to metal contact with the block, with the upper face .1 to .15mm above the block achieved without selective assembly with the gaskets making up the difference. A relatively high (85 lb/ft (115Nm)) torque loading of the 12mm thread rolled studs was required and the assembly was proved to bed running a type 906 engine (the racing version) on a series of thermal cycles, cold idle to max mep for five minutes the idle and cool, repeat cycle. There was full cold water circulation at idle, no water circulation at full throttle. The gasket stood up to 500 of these cycles better than the factory occupants. Subsequent road running showed an electrolyte corrosion problem affecting the gasket due to presence of many dissimilar metals and acidic antifreeze, this was cured using a stainless gasket and inhibited coolant. There was also a hot clearance problem with the main bearings in an all aluminum crankcase, which showed as a low idle oil pressure, which although not in itself harmful, was felt to have bad sales psychology. The Layshaft drove the distributer and coaxial oil pump, thus couldn’t be speeded up, and an unnecessarily large oil pump had to be fitted. One main bearing panel cracked from the fillet radius adjacent to one of the centre main bearing suds to the bearing bore. The fillet was deleted and the section increased. Road testing showed high oil consumption, around 900km/l due to high liner distortion probably caused by the necessary high clamping loads, so the liner wall thickness was increased. A great deal of piston ring, piston profile development took place before a combination which gave scuff free production customer running in with acceptable low oil consumption.
Some of these engines were supplied to Jensen for test in Healey prototypes, which involved a modified sump and exhaust system to suit the chassis layout. These early engines had cut and welded aluminum sumps and fabricated steel tube exhaust manifolds and performed reliably and well a series of exhaustive testes in prototype Healey’s. Transmission vibration problems led to the bean stiffness of the engine gearbox becoming suspect, the two lugs (see Appendix 357.1 Fig 357.8) were added to the crankcase lower half. Subsequent tests showed the beam stiffness was 30 percent greater than contemporary engines of similar size or output.
These encouraging results lead to near disaster. The Healey program was advanced cutting down one end of the pre-production program, which appeared an acceptable risk, but delays in working up the machining facility on production die castings combined with the 1972 fuel crisis, drastically shortened the pre-production validation program with inevitable results. When production engines were fitted to pre production cars, tow serious oil system problems appeared (both above 6000 rpm) high speed continental cruising moved oil via the breather system into the air box and the oil pressure declined to a dangerous level. As usual prototypes were found to be at the favorable end of the tolerance range, correctly produced one piece cast sumps did not cool as effectively as development lash ups and the time taken to bed in piston rings varied. A short-term solution was made with an external velocity dropping oil separator in the breather system which returned the oil carried out by the crank case blow by to the sump. A longer-term solution was achieved by forming the housing for the rubber lip type crankshaft seal into an oil separator chamber. The initial rope seal for the crankshaft proved to be prone to assembly errors. The oil pressure problems were partly due to oil being held up in the cam boxes by miss-matched drain holes, casting flash, sticking and dimensionally incorrect relief vales, which also showed up as long delays in reaching satisfactory oil pressure on starting. Urgent remedial action to remove the oil from the cam boxes led to exhaust camshafts scuffing due to lack of lubrication on starting. Cured by phosphating the cams alone. This urgent final development took place while commissioning the machining facility and training operators. It was feared that emission development would absorb too much power. By reading across from the ‘Big Valve’ version of the 1560cc twin cam, (which used the same tappets, valve springs, etc.) and the racing kit of parts, a reliable, drivable 178hp (133kNm/s) engine with 206Nm torque was developed, using slightly larger valves, different camshafts and tuning.
As the emissions program developed it became clear all the features of this version of the engine would not be immediately required. It was also decided to use one basic engine for all purposes and change only distributor, inlet manifold and carburetors, to meet different market area emission requirements. Finally, a great deal of detail development also went into reducing test bed and vehicle assembly line rejects from the 20 per cent after the first six month production to the present 2 per cent. The most serious problem, apart from oil leaks, was jumped timing belts on cold star, a number of which were proved to be caused by production debris. But a increase in cam-belt running clearance of .025mm provided a permanent cure and made a significant reduction in noise level. Other development centered on reducing machining and assembly time. For example, it took six man hours to produce sixteen satisfactory sealing valves when production began, it now takes 0.5 man hours to produce sixteen highly effective seats.
5. Emission Development
It was believed that the race bred techniques of controlling the rate of burning by applying ‘squish’ to a compact combustion chamber, could produce sufficiently complete combustion, at the low speeds required by the Federal test cycle, providing that the ‘squish’ areas were not so large as to quench or over cool the charge to the degree that patches of partially burnt charge were released into the exhaust.
The four valve layout was invaluable as it permitted a compact combustion chamber, with a centrally located extended nose sparking plug, igniting the charge as near its centre of volume as possible. Further advantages of four valves were that adequate filling and exhausting could easily be achieved without wild cams and/or abnormal overlap and much better control of gas temperature on the exhaust side was possible, with calibration of the water flow around the two exhaust seats.
It had been accepted that the feature limiting the power of the emission engine, would be the volume of air passed by the available Stromberg 1.75CID Carburetor and would probably be just over 140 hp (104 kNm/s). To give the engine the best possible breathing the large valve layout was standardized.
The first engine used for emission testing ran quite happily on 91RON unleaded fuel, with less than .025mm recession of the valve/valve seat per 100h, but audibly protested on the road at low rpm part throttle. Since proposed Californian taxation would penalize engines over 8.5:1 and there was power to spare, a bowl, 3.8 mm deep was machined form the crown of the pistons to give 8.4:1.
Serious testing began with this version, with the larger valves and 8.4:1 CR which gave about 150 hp (112 kNm/s) with the four fixed choke arrangement and 144 hp (107 kNm/s) with the Stromberg emission carburetors and water heated manifold.
The emission levels (grms per mile) were: HC 3.0, CO 9.4, NOx 1.2. Since the 1974 HC limit would be 3.4 grm permile there was not enough margin for safety. The valve overlap was reduced form 52’ to 42’, which reduced the maximum output to 142 bhp (Fig 357.4a) although small, in time, measured as lift area, it was a 50 per cent reduction in lift area as 357.5 shows. This change gave emission levels (grs/mile) of: HC 1.46, CO 12.5, NOx 1.5 and cost between 2 and 4 hp.
Time was running out, so this specification was adapted form the 1973 Federal engine, which was certified in the Healey with deterioration factors applied at: HC 2.8, CO 22.0, NOx 2.8.
These were, of course, production engines with pre production tolerances, and meant that despite the engine being the first certified for 1973 on a clean air package, further work would be required for 1974. Combined attention detail, particularly in the Quality Control field, such as accurately controlling water flow around the valve seats by ensuring accurate matching of water passages, block-gasket-head, have reduced the scatter between good and bad engines so that all engines are below the former ’good’ level.
For 1974 the ignition retard sensing in now modulated from the manifold, instead of throttle edge drilling, the valve overlap has reverted to the original wide setting as better oil control and carburetor quality control have reduced HC. There has been a positive improvement in exhaust valve and seat cooling, by better quality control as described. Lighter carburetor bypass valves were also effective.
At the time of writing, it appears that the 1974-certification levels will be, in gms mile, HC 2.3, CO 2.9, NOx 1.40.
This does not give a very comfortable margin of safety for production scatter and is therefore receiving maximum development priority.
Unsuccessful experiments so far have included:
(a) Increased water circulation around the inlet manifold.
(b) Oil jet from connecting rod on the underside of piston crown
(c) Move valve overlap still, NOx reduced HC very nearly off scale.
(d) Thermostat de rated, 30 per cent reduction on NOx, car failed demist test.
The original engine concept, decided upon before the more stringent specification requirements were enacted, has proven successful in providing a reliable, high performance, low pollution engine, free from complex regulation beating accessories. It is in fact remarkable and fortunate how effective a number of decisions dictated by extreme expediency have combined to produce an acceptable solution. In retrospect, the same emission results could probably have been achieved with a three-valve engine using tow exhaust valves.
The Author wishes to thank the Chairman and his fellow Directors of Lotus Cars Ltd., not only for the permission to produce this paper, but also for their patients and support while the project was brought to fruition. Also Mr. R. B. Burr who assisted in the preparation and was Chief Designer at the time the engine was originally designed, and Mr. S. G. Williams, Powertrain Engineering Manager of Lotus Cars, for his assistance in preparing the paper.
Two Litre four cylinder in-line engine inclined at 45’. Aluminum cylinder head and cylinder block with cast iron wet liners. Belt driven double overhead camshafts with four valves per cylinder. Two basic specifications: one for Europe one for North America.
Capacity 120.5 ci 1973cc
Bore 3.751 in 95.28mm
Stroke 2.716 in 69.24mm
Compression ratio 8.4 or 9.5
Dry Weight (including alternator
starter, but no clutch) 275 lb 125 kg
Cylinder block and main bearing housing
Aluminum alloy LM 25 WP gravity die-cast, pent roof combustion chambers with four valves per cylinder and central 14mm spark plug. Included angle between valves 38’. Brico sintered valve seats, cast iron valve guides, 21 4NS one piece exhaust valves, EN 50 inlet valves and double valve springs. Water takeoff facilities on both sides and both ends of head. Steel/asbestos/stainless steel cylinder head gasket.
Separate but identical aluminum allow camshaft housings for inlet and exhaust. Bucket tappets running directly in the aluminum. Camshafts running directly in the aluminum in five integral bearings. Valve adjustment by biscuit shim. Drive by Powergrip toothed belt 9.5mm pitch, 25mm wide.
Valve Timing Inlet opens 25 21 BTDC
(nominal Inlet closes 65 71 ABDC
.25 mm Exhaust opens 65 71 BBDC
ramp) Exhaust Closes 25 21 ATDC
Different cam profiles for Europe and North America.
Aluminum alloy carrying jackshaft, distributor, concentric oil pump, and oil filter. Also carries location for alternator and oil cooler take off facilities.
Idler pulley mounted on front of block. Manually adjustable by eccentric cam.
Hepworth and Grandage solid skirt pistons. Plain top ring, Stepped second compression ring. SE oil control ring. Fully floating 25mm gudgeon pin retained with square section circlips. Valve clearance recesses in crown, which is dished for low compression engine and flat for high compression engine.
Material En 16 T. Centers 139.75mm. Split pin dowels.
SG Iron (SNG 37/2) casting. Single drillings to big ends.
Vandervell VP2 copper/lead
Aluminum alloy. Flat face to bearing housing