Ken Helmick Feedback

From Open Source Ecology
Jump to: navigation, search


We haven't been introduced, so allow me to take some liberties. My name is Ken Helmick and I'm a resident of southeast Michigan. I am a recipient of e-mails to and from SACA President Tom Kimmel and am therefore assuming he'd like me to add my thoughts. Please excuse any abruptness that may occur in this mail, I'm in a lull between assignments (which may end any moment) and therefore may unintentionally come off as brusque in my hurry to finish before that occurs. For much the same reasons, this message may tend to be a bit unorganized.

The problem of designing and building a steam engine involves many parameters with initial decisions having huge influence on final design. Higher operating pressures and temperatures lead to potentially higher efficiency, but the design must be optimized to realize this potential else the energy and initial cost expended to generate this highly energetic steam is wasted.

Generally speaking, a uniflow engine has the potential of higher efficiency, but in order to achieve this the clearance and cutoff are necessarily tightly controlled and (preferably) the exhaust is subatmospheric. A counterflow engine has the relative advantage of being more easily operated across a wider variety of steam inlet conditions at the cost of higher engine complexity and potentially lower efficiency.

The issue of cylinder liners was bought up. Wrapping a sleeve and welding is not likely to yield a satisfactory engine for a whole variety of reasons I won't attempt to describe in detail. The methods I would employ would include:

  • Fabricate wooden pattern and core box, sand cast iron cylinder, machine. I actually have such a pattern, core box and cast cylinder in my basement.
  • Purchase engine sleeves off the shelf--one example of a manufacturer capable of supplying both finished and 'raw' centrifugally cast sleeves is LA Sleeve Co.
  • Purchase already honed hydraulic cylinder tubes and fabricate the engine around those. One of many suppliers is
  • Convert an existing IC engine or extensively utilize IC engine components. Some motorcycle engines have replacable cylinders (notably V twins) and could be readily modified, as can the original VW Beetle Boxer engine which can be purchased in any variety of original or aftermarket configurations and level of assembly.

Use of the centrifugally cast sleeve would require an external retaining cylinder which could be either cast or machined from stock metals. Honed hydraulic cylinders could be used the same way or, in the case of cylinders with heavier side walls, it may be possible to join upper and lower flanges as well as exhaust manifolds directly to the cylinder. If welding is employed, I'd likely stress-relieve the assembly and then re-hone lightly, just in case there is any slight thermal induced distortion to the bore.

One problem with bump valves is that of valve springs. If higher temperatures are contemplated, special springs must be wound from superalloys and correctly heat treated. This is well beyond the home machinist and probably some more pedestrian heat treat vendors. Conventional springs will rapidly fail under higher steam conditions. Another problem with bump valves is mass, the heavier they are, the more potential damage due to impact, this is why Jay Carter invented the light weight valve. Too many bump valve engines had very short life between failures.

At first blush, bump valves shouldn't work. The valve is open just as long before TDC as after and the combined effects of compression and steam pressure admission just before TDC should rob as almost as much power as steam admission and expansion after TDC. This would tend to imply that steam inertia advantageously provides for asymmetrical flow around TDC. The corollary is that the acceleration effects are tied to the amount of time the valve is open and, therefore, RPM must come into play very significantly. I'd postulate that depending on size, inlet geometry and pressure differential the engine rpm has an effect on how well the bump valve engine performs with performance and efficiency degrading both above and below some particular rpm 'sweet spot'.

Asymmetrical valve events for bump valves can be achieved by either staggering the cylinder according to the DeSaxe principle so that the cylinder axis does not intercept the crank axis or by offsetting the piston wrist pin. The second method is very common in modern IC engines although the wrist pin offset is done primarily to help the piston distribute the peak cylinder pressure against the side wall over a longer time period and thus reduce wear. In any case, offset should improve the bump valve engine operation but will lead to deteriorated performance (or even non operation) in reverse.

Besides the thought that bump valve engines probably are more efficient at some moderate rpm, there are other reasons for operating at higher speeds than traditional steam engines. Blowby around the piston rings is a significant loss of power and efficiency. This is proportional to both the mean effective pressure and the residence time in the cylinder. For high steam pressures, a very short cutoff lowers MEP to a point where blowby is less of a problem. Higher rpm provides less time for the blowby to occur. In addition, the short cutoff (high expansion) of most uniflow engines and associated low MEP reduces the amount of power generated per unit of cylinder volume. All things being equal, power goes up wirh rpm, so this is a way to gain back power lost by pursuing more efficient expansion. Fluid lubricated bearings employ hydrodynamic suspension (Langmuir theory of lubrication) to reduce friction and wear. The strength of the suspension film rises with shaft velocity, and so does the load the shaft can bear without bearing failure. Typically, lugging wears out more car engines than racing.

There is a potential alternative to the bump valve that is otherwise pretty similar in overall operating characteristics. Basically, this is a smaller piston valve mounted coaxially with, and upon, the engine piston. This piston valve opens to admit steam only around TDC and is otherwise closed...the same condition a bump valve attains. The disadvantage is more friction, extra rings or seals and the need for more precision in manufacture. The advantages are no springs to weaken from temperature and no hard impact leading to valve failure.

I would always enclose the crankcase of any higher rpm engine, smaller such engines are traditionally referred to as 'splash lubricated' for very good reasons.

I agree with Tom concerning piston rings, there is an incredible variety available in almost any size, configuration and material imaginable and this is about the last reason I would select a given engine diameter. I have chosen a 4 inch bore because it is about the largest diameter used in modern passenger automobile practice, not because of the rings. Possibly the supplier with the best selection is a local company, check their pdf download for diametric applications:

I don't know that I see an issue with an electric starter. The electric power supply can be incredibly simple. Older General Motors alternators are widely available and they have an integral voltage regulator. Simply connect the alternator to a 12 battery with standard automotive cables and a self regulating system is in place, explaining why these units are often seen in hotrods built up from other makes of cars. This also allows for niceties like electrical steam plant controls and a light to see what is going on in the dark.

A uniflow exhaust manifold could be very, very simple, depending on the cylinder. If using something like a honed heavy wall hydraulic tube or a cylinder with a liner, I'd consider haunting the steel supply houses and finding a piece of larger rectangular tube, say maybe something like 2 x 6 inches. Weld a cap on one open end of the tube and a flange on the other, bore a hole in the face equal to the cylinder OD, slide over the cylinder and lightly weld.

The crankshaft is the heart of the engine. All the loads focus on the crank and it is subject to compression, torsion, tension and shear loads as well as cyclic fatigue and potential damage from localized heating; proper design and fabrication is necessary to minimize destructive inertial unbalance loads, misalignment and bending. People have been making cranks for a long time, and almost anyone can do so, building a crank capable of trouble-free operation at high speed and power for long periods of time is another matter.

The average home machinist is typically not equipped for (nor capable of) building a multiple throw, one piece crankshaft capable of turning high rpm and decent torque for extended periods; this isn't an insult but reflects the high degree of specialization needed to produce such an item. The most certain and economical means of obtaining such an item are to buy a crank used in a mass produced engine.

If that isn't feasible, one route to consider is the built-up crank. These are found in many smaller engines such as those used in ATVs and snowmobiles. Since the crank is assembled, it is possible to use ball bearings in the connecting rods.

If a one piece crank is desired, the best route to go would be to either cast a rough out of a high grade of iron (still used in some V-8 engines) and machine it leaving the pins and mains a bit over sized. Another option would be turning the crank a bit over-sized from billet steel. Many engine rebuild shops have the equipment to accurately grind mains and pins (the pins are the hard part) as well as microsize (polish) to final dimension. Manufactured cranks tend to be fillet rolled to increase toughness, but this won't be feasible, so filleted pins would be needed along with matching bearings. Such bearings are stocked by racing crankshaft builders. Before having a rebuild shop grind and polish the crank, it should ideally be stress relieved and possibly nitride hardened; something available from job shops. The home machinist can balance a crank with a single throw (assuming he remembered to add counterweights to the design and fully understands the issues involved) but multiple pin cranks can only be balanced in a shop with a dynamic balancing machine---many hot rod shops have such capability

Even production cranks will need to be rebalanced if changes are made to the pistons, connecting rods and so on UNLESS the crank is of a symmetrical design. Typically, this would be radials, inline and boxer 4 cylinders, inline 6s, inline 8s and V-12 engines.

Hope these initial thoughts are of some help.