(SACA Board member Tim Nye, learning that I was planning a visit to the University of Michigan’s Buhr Shelving Facility, asked me to scan the following article.  It is an English translation of a Czech article, right at the end of WW2. It is fascinating in that the Czech engineers were running a stationary steam engine without cylinder lubrication, instead using carbon rings to seal the cylinders. Ed.)  
 
HIGH SPEED STEAM ENGINE WITHOUT LUBRICATION 
By H. Frank and J. Rais. (From VDI Seitschrift, Vol. 89, Nos. 5/6, February 3rd, 1945, pp.6466). 
 
(Translation appearing in The Engineer’s Digest, October, 1945   Vol. 2, No. 10.) 
 
A STEAM Engine to operate without cylinder lubrication has been designed and built by the Skoda Works (Czechoslovakia – Ed.) for a special purpose, where oil-free exhaust steam and highest economy at even partial loads are vital requirements. The engine was built for an output of 440 h.p., with live steam of 43 kg. per sq. cm. pressure and 380 deg. C. temperature, the back pressure being 7-9 kg. per sq.cm. 
 
(steam at 612 lbs. per sq. inch, 716 degrees F and 100-128 psi back pressure. This engine was briefly mentioned by Stan Jakuba in the April/May/June 2000 issue of the Bulletin.  He noted it had a water rate of 18.3 pounds per horsepower-hour. This poor water rate is undoubtedly due to the low expansion ratio responsible for the high back pressure. It’s almost certain this industrial engine was designed to produce a high back pressure so that the exhaust steam could be used in other industrial processes.  Ed.)
 
Aiming to secure oil-free exhaust from a reciprocator, the Skoda Works had carried out extensive experiments with regard to the use of carbon seals.  However, these experiments, conducted since 1935, had been confined to considerably lower pressures. Thus, when faced with the present task, a new approach to the problem had to be made. The new series of tests were concerned with finding the most suitable type of carbon material, most suitable, that is, with regard to mechanical strength, thermal expansion, wear, and ability to withstand high operating temperatures. The most suitable material for the piston and the piston rod had also to be ascertained. Furthermore, an existing engine was equipped with carbon seals in order to measure leakage and other data. Carbon is adversely affected by the presence of oil, a sticky mass being formed which exerts an abrasive action upon the piston rings and wears them down with great rapidity. Particular attention had therefore to be paid to preventing lubricating oil from the crank case, from reaching the carbon seals. In this respect the provision of duplex scraper rings with inter-drainage proved most helpful. These rings are now made of sintered iron in place of bronze, as hitherto used. 
 
(This section can be interpreted to mean that the piston was affixed to a piston rod which, in turn, was reciprocated by a crosshead. It’s impossible to speculate on whether it was single or   In this manner the crankshaft, connecting rod and crosshead could all be oil lubricated without fear of introducing oil to the carbon rings. Since the piston has no appreciable side loads, likely not even contacting the cylinder wall, no lubrication would be required for anything but the piston rings.  Ed.) 
 
Carbon is a very good material for sealing purposes and by no means inferior to a metallic seal provided the rubbing parts to be sealed are as hard as possible and have a mirror-type surface finish. Should this not be the case, carbon particles will lodge in any existing pores and wear of the carbon rings will be inevitable. 
 
A horizontal engine design had to be ruled out for reasons of cost, and a vertical design had therefore to be resorted to. A multi-cylinder design had to be chosen, since large size carbon seals are net feasible because of danger of breakage. It was, moreover, necessary to decide upon a single-acting engine type in order to eliminate the stuffing boxes. The specified engine output called for a six-cylinder design with a bore of 
199 mm. and a stroke of 200 mm., the engine speed being 750 r.p.m. This results in a mean piston speed of 5 m. per second, which is a high value hitherto rarely used in steam reciprocator practice.  
 
(bore of 7.835 inches, stroke of 7.874 inches at a speed of 984 feet per minute, similar to that of some Doble engines – Ed.) ​

(The magazine illustrations weren’t sometimes too clear.  The magazine was almost 70 years old and it’s possible that British wartime rationing didn’t help as regards paper and ink quality.  
The drawings have been either modified or replaced as possible. Absolute reliance on the images should not be taken as they are best interpretation.  Ed.) 
 
Upon removal of the cylinder head, the carbon rings (a), as shown in Fig. 1, are slipped over the piston (d) together with the ring carriers (b). Circular leaf springs (e) serve to press the carbon rings lightly against the wall (f) of the cylinder. The piston crown is connected with the piston rod by means of the bolts (h). As these bolts exhibited fissures after short time in service, they were subsequently replaced by longer ones. The method employed for sealing the valve spindles is outlined in Fig. 2. Here it will be seen that the carbon bush (b) serves both as guide for the spindle (a) and as first stage gland, the steam pressure being throttled to a certain extent by the labyrinth-action of the grooves provided in the carbon bush. Sealing of this carbon bush against the valve body is effected at top and bottom of the bush by means of the elastic copper packings (c), the required degree of compression being exerted by the threaded gland (d). ​

The stuffing box proper is constituted by the ring carriers (e)containing the carbon rings (f). Each of the latter is composed of three circular segments, the segments of adjacent rings being staggered by 60 deg. (Detailed section A-B in Fig. 2). The carbon segments are pressed against the valve spindle by means of the insular leaf springs (g). Furthermore, locating pins are provided to keep the carbon segments from rotating. In order to prevent the leakage of steam to atmosphere, distance piece (h) is inserted below the two topmost rings, this annular distance piece connecting with the back-pressure steam space by means of the duct (m). The two topmost carbon rings therefore act only as seals against the back pressure of the engine. In this way steam loss is kept to a minimum.  
​
The rings (i) and (k) act as oil wipers, with the leakage oil being led away through the opening (l).  
The mutual angular disposition of the six cranks of the engine crankshaft is indicated in Fig. 3. The connecting rods are forged and milled in order to keep their weight to a minimum. The big end bearings are made in cast steel halves, while the crosshead bearings are of the box type. The light metal crossheads have wrist pins of special bronze. The crank case is a single welded piece in order to avoid bolting the crank shaft bearings; the latter are bolted to pedestals enabling the withdrawal of the crankshaft to one side without lifting it. The cylindrical crosshead guides are made in halves and can be easily removed together with the crossheads through the openings provided in the engine casing. In the original engine the piston rods and certain other parts were made of stainless steel, but this precaution was later found to be unnecessary.  ​

The employment of a mechanical governor had to be ruled out because of the high engine speed and the large controlling forces required, and a pressure oil governor system was therefore adopted, as shown in Fig. 4. The pressure oil pump as well as the gear type lubricating oil pump are driven from the engine shaft.  
​
Referring to Fig. 4 it will be seen that the small fly-ball governor (b) acts upon the pilot valve (c) of the slave cylinder (d). Valve control is effected in a novel manner, there being a cut-off control lay-shaft in addition to the main lay-shaft. The admission valves (e) are controlled by the cams (g) and the exhaust valves by the cams (h). The main lay-shaft revolves at constant speed, while the angular position of the cut-off layshaft relative to the main lay-shaft can be adjusted by shifting the pinion (i) by the movement of the piston of slave cylinder (c). Electrical remote engine speed control is provided. The engine is equipped with lubricating oil and pressure oil coolers, oil filters, and an emergency governor with oil hydraulic operation of the throttle valve. A manually operated pressure oil pump to set up the required oil pressure for starting is also provided. Provision for the absorption of thermal  expansion is made at all vital points, including the steam piping. ​

The steam supply line from the boiler was laid out with a minimum of bends, a distributing header being provided close to the cylinders. From observations made at points of the live steam connections close to the admission valves, it was found that there prevailed a weak steam  oscillation of a frequency corresponding to the rotational speed of the engine. In addition to this, higher harmonics and a certain amount of non-harmonics were also found to prevail. The most suitable means for the suppression of these oscillations has not been found as yet. As the compression amounts to about two-thirds of the admission pressure, there also occurred a small amount of vibration of the admission valves, but this could be eliminated by changing the valve springs.  
Compared to a slow speed poppet-valve engine of equal power at 125 r.p.m., the present engine affords a saving in weight of some 60 per cent. At the end of August,1944, the engine had been in service for about 8,000 hours without showing any marked wear of the carbon rings. The amount of cylinder lubricating oil saved during this period amounts to more than 5,000 kg. The guaranteed specific steam consumption was 11.2 kg. per kW.-hr. (18.4 lbs. per h.p.-hr. -  Ed.) with a tolerance of +5 per cent, with a load of 410 h.p. This compares with a measured consumption of 11.95 kg. per kW. hr.,(19.6 lbs. per h.p.-hr. – Ed.) that is, +6.7 per cent in excess of the guaranteed consumption. From diagrams taken at a later date, it was, however, found that two cylinders showed insufficient compression owing to faulty adjustment of the valve control. It is therefore confidently expected that after proper adjustment has been made the guaranteed steam rate will be met without tolerance. The indicator diagrams reproduced in Figs. 5 and 6 were taken with an electrical indicator. They clearly show the resonant oscillations occurring during the admission period. ​

(A couple of notes seem worth mention. It mentions that the cylinder walls should be quite hard and as smooth as possible. This seems reasonable as graphite is soft and would wear away against an even mildly abrasive surface. Industrial hard chrome (not the stuff put on cars and motorcycles for the sake of being shiny) sounds like it might fit the bill, if polished.  To be clear, this would be the kind of chrome plated onto injection molds to reduce wear … and other such heavy-duty work. 
Hopefully the article has been of interest.  It would seem to follow that, if this approach worked on engine cylinders, it might also be valid for piston valves.  It becomes more debatable for slide valves given the pressure applied to the sealing surfaces.
  
After having slaved over a hot keyboard to put the article into a computer, and reworking some drawings, Tim Nye pointed out that that I had missed the fact that Josef Rais, one of the authors, had previously taken out patents on carbon piston rings and seal.  The US patent, which is probably the most easily accessed, was applied for in March of 1934 (originally April 21, 1933 in Czechoslovakia) and was granted on April 4, 1939. It carries the title “Packing  Ring for Sealing the Pistons and Piston Rods and was assigned to the Skoda Works in Pilsen, Prague, Czechoslovakia. I have to thank Tim for this observation, it brings the article into much clearer focus. Rais makes the case for his patent, as follows: 
“This invention relates to a new and improved type of packing ring for sealing the pistons and piston rods of reciprocating engines and machines using working cylinders and reciprocating pistons. The known constructions of engines and machines of this type suffer from the disadvantage that the packing members must of necessity be lubricated with some special lubricant supplied from without. This requirement results in a considerable increase in the cost of running, and expensive means are necessary for reducing the quantity of lubricating oiI to the minimum. In steam engines, the lubricating oil contaminates the exhaust steam, so that the latter cans not be used for certain purposes, since it contains excessively large quantities of oil. 

In the case of compressors, the lubricating oil can be caused to ignite. In consequence of the increased air temperature, thereby resulting in an explosion and hence the destruction of the entire machine. (When working with 5,000 psi air compressors, while in the Navy, we used special vegetable based oils for cylinder lubrication specifically to avoid these explosions.  Ed.) Furthermore, the coolers are apt to become fouled by the lubricating oil, so that a special oil eliminating device becomes necessary. In the case of compressors in which a lubricant other than oil is employed, the pistons must be provided with soft packing means, for example cups, which are, however, comparatively short lived. In the case of refrigerating machines also, the lubrication with oil, glycerine, and other lubricants often gives rise to troublesome disturbances. Similar inconveniences due to the necessity for lubrication occur in the case of internal combustion engines also.  

The present invention overcomes these drawbacks by virtue of the arrangement that the packing members of the pistons or piston rods, or of both, consist of rings of carbon, graphite, or similar frictionless self-lubricating material, which rings are so resiliently mounted relatively to spring metallic supporting or holding means, that the necessary radial resiliency is provided without any independent lubrication whatever.  

These carbon rings are inserted in metal holding bushes, both the ring of self-lubricating matter's and also the resilient metal holding means being cut through by a single cut, for the purpose of enabling the entire packing member to be resilient radially.  

The packing ring according to the invention is rendered particularly well able to withstand the strains and stresses to which it is subjected by virtue of the arrangement that the ring consists of two coaxial independent rings loosely mounted one upon the other.”   

Rais notes: “In the hitherto known constructions of packing rings of carbon, graphite, or similar material the ring is either divided into a plurality of segments or else is rigidly connected to a metallic base. The first mentioned principle of construction does not ensure a perfect seal nor a uniform pressure of the packing ring against the co-operating surface, whereas with the second principle of construction the Independent stressing of the carbon packing ring is prevented, and in addition detrimental stresses are transmitted from the underlying base to the carbon ring whereby either the connection between the carbon ring and the metallic underlying base or else the carbon ring itself is damaged.  

These disadvantages are overcome by the present invention, by virtue of the arrangement that the metallic ring is loosely inserted in the carbon ring, with possible locking against lateral displacement. In accordance with the invention, the packing ring may for this purpose consist of two coaxial and independent rings loosely mounted one upon the other. These two rings are mounted one upon the other without any rigid interconnection whatever, with the result that the resilient action of the metallic ring is uniformly transmitted to the carbon packing ring. The sealing action of a packing ring constructed in accordance with the invention can be increased or reduced as required in the various cases of utilizing the rings in machines by choosing suitable metallic resilient material for the production of the resilient metallic ring. The seating surface between the two rings can be cylindrical; but may with advantage be provided with one or more grooves of any desired cone figuration for the purpose of preventing axial displacement. Packing rings constructed in this manner permit of radial movement of the ring performing the packing function. In this case, however, if the grooves be of dovetail or other shape, it is necessary to provide such connection between carbon ring and metallic ring with an amount of play which admits of displacement within limits permitting of perfectly uniform resilience of the packing ring. 

The resilient packing ring according to the invention is cut through in a normal manner at one point on its periphery, the cut in the packing ring and that in the metallic ring but either coincident or in offset relation to each other.” 
​
The following illustrations are all derived from the patent, while the abbreviations in the table replace the number designations used to describe patent features (on the theory that it might be easier to remember MR stands for “metallic ring” rather than the number 13, or whatever). ​

Figure1 shows three rings installed in the piston while Fig. 2 describes the rings themselves. ​

Figures 9 and 10 consist of packing assemblies holding multiple carbon rings. ​

Figures 11 through 13 show pistons built up in segments, which serve to entrap the packing elements. The bushings are either a springy retainer or are backed by leaf or helical springs. The stops distribute the spring load against the bearing surface after the carbon rings have reached the worn in state. Thus, the springs are actually a temporary expedient used to break in the packing element. ​

Generally speaking, the above embodiments are used with split, built-up pistons. Figures 21 and 22 are designed to be installed into the groove of a solid piston.  Leaf springs underneath force the carbon ring against the cylinder walls while the stops are pins driven into holes drilled into the piston grooves from either the top or bottom of the piston.  The rings are made in halves and use tongue and groove joints to promote tightness. ​

Rais noted: “In accordance with the invention the packing ring may be constructed in such a manner that its resiliency in a radial sense is effected by metallic means provided beneath the ring.”  This implies that he understood the metallic ring would supply the resiliency required to hold the carbon packing in contact with the sealing surface but did not rule out the natural resiliency of the graphite itself. 
​  
Today, we might have a greater range of options.  At least two companies manufacture carbon materials engineered with an array of properties.  Graphitar products are manufactured by US Graphite in Saginaw, Michigan while Superior Graphite is headquartered in Chicago, Illinois.  
Superior Graphite has been in business since 1917 while US Graphite was a spinoff of the Wickes lumber business and was founded in 1895 to exploit then-recent graphite discoveries in Mexico. 
The following comes from the Superior Graphite website found at:  http://graphitescarbons.superiorgraphite.com/Asset/RGC-Brochure.pdf ​

(RGC stands for ‘Resilient Graphitic Composites’, Ed.) ​

 
                 The Development of a Steam Powered Motorcycle
​                                            PART ONE
The following article covers the basic design philosophy for an attempt at the land speed record for a steam powered motorcycle and perhaps the outright record for any steam powered vehicle.

As a starting place for a steam powered land speed record, a motorcycle is a terrible idea. The restricted space envelope and poor aero-dynamic characteristics make it the worst possible concept. However, not to be daunted that is what we chose. The decisive factor was  because we had this small Bower and Bell engine, which was an Abner Doble inspired design, it is too small for a car, but deserved to be used in something. Our love of motorbikes and the idea of a steam powered motorbike is more appealing to us than the ultimate record itself, we have used a Suzuki Hayabusa as the donor vehicle.

The basic formula pertaining to drag is: Drag = Cd x (1/2 x density of air x velocity2 x Area), where Cd is the co-efficient of drag and Area is the projected frontal area. As we can see, drag increases by the squared function of speed. The two most important factors that we can control in design space is frontal area and the Cd.

A motorbike has a relatively small frontal area, which is good,. However, even fully faired racing motorbikes struggle to achieve better than a Cd of 0.6. This compares to 0.3 for a modern road car, a Streamliner can have Cd as low as 0.12.​From the RSR Bonneville Aero-Horsepower & Drag Loss Calculator: available online at: - 
https://www.rbracing-rsr.com/aerohpcalc.html 

Coefficient of Drag: Street, faired motorcycles are notoriously inefficient aero-devices with Coefficient of Drag (Cd) figures in the 0.6 range. For example, a Suzuki Hayabusa has a Cd of .561 whereas a Kawasaki ZX-12 has a Cd of 0.603. Modern cars often have paid close attention to aerodynamics and may have Cd figures of 0.3. Streamliners may have Cd figures of under 0.2, perhaps as low as 0.15 or in some cases figures of 0.10 have been achieved.
 
Frontal Area: Reducing frontal area is key to going fast as the horsepower requirements go up exponentially as you push that "barn door" through the air. You'll need a close approximation of your vehicle's frontal area in square feet to make this calculator entry. A Suzuki Hayabusa has a frontal area of 6.01 square feet. A Kawasaki ZX-12 has a frontal area of 6.09 square feet. Some streamliners like the Lambky Vincent have only 4 square feet frontal area. 
​
Vehicle Weight: On shorter courses with asphalt surfaces and good traction weight is more of an issue than it is at Bonneville, where weight can aid traction on the slippery salt surface. Short courses are more of a drag race and accelerating extra mass is not a good idea. At Bonneville the big dogs will be on the long course with over six miles of salt with the clocks at the 2, 4, and 6-mile markers, so weight is not nearly as much of an issue.

                                 Figure 1: An example of a record setting Streamliner at Bonneville ​

                                      ​Figure 2: At Elvington airfield UK setting 80mph run

A mechanical water pump is fitted to the end of the free-wheeling hub output shaft. This delivers all of the water necessary. The capacity of the water pump is approximately 200% of theoretical demand, ensuring the ability to rapidly increase the water level within the steam generator if steam temperature is too high.
​The water pump is driven by an eccentric, which can be changed to provide different water delivery, should testing reveal changes are required. Water must be added to the steam generator at start up by external means, as no electrical or hand pump is provided on the bike.
 
We have chosen to run the steam generator at between 900 and 950 Fahrenheit. The steam lubricating oil limit is 1000F. This increase in temperature results in reduced water consumption for a given power. Abner Doble typically used 850lb/hr at 850F as the requirement for a 100Hp engine. This aligns with the water consumption predicted by my spreadsheet. 
                                                         Steam Generator 
The steam generator has been designed to output sufficient steam for 96Hp, which is 750lb/hr at 950F. The basis for all fluid velocities and required surface area has been calculated from the work of Abner Doble, including the required combustion volume. The steam generator has ~430 feet of 8mm OD tube for the water coils, 180 feet of 12mm OD tube for the steaming coils, 30 feet of ½ OD tube for the Normaliser and 30 feet of ½ OD tube for the super heater. The water and steaming coils are carbon steel for improved heat transfer properties. This minimizes the required tubing and the ​​Normaliser and Superheater coils are 316L stainless steel to withstand the higher temperatures. The total weight of tubing is 120lbs and the internal volume is 0.3 cubic feet. If one considers that the required output of 727lb/hr equals 12lb/min and the internal volume is sufficient for ~9.5lb of water/steam, then simplistically the approximate time that a molecule of water entering at the inlet to the generator takes to subsequently exit the generator at 950F is (9.5/12) *60 = 47.5 seconds. At full power the velocity in the water tubes is 9ft/ sec, in the steaming coils 50ft/sec and in the superheater 75 ft/sec. Therefore, notionally the water spends 48 seconds in the water coils, 3.6 seconds in the steaming coils and 0.8 of a second in the superheater. 
​
This might seem incredibly fast. However, comparing to Abner Doble’s design for a 100Hp generator producing 850lb/hr of steam at 1000psi, (steam at this temperature and pressure has a density of ~1.39lb/cubic foot). 

Producing 850/3600 = 0.236lb/sec. 
​
The super heater and normaliser coils are 3/4inch OD, 9/16inch ID and 39 feet total length. The cross-sectional area is therefore 0.2485inch2. The volume in the superheater is (39*0.2485/144) = 0.0673feet3, so contains 1.39*0.0673 = 0.0936lbs of steam. Therefore, to produce 0.235lb per second, requires the superheater to be emptied 0.235/0.0936 = 2.5 times per second, 2.5 times 39 is 98 feet per second velocity!

                                           Figure 4: Steam generator casing ​

                                                         Figure 5: Schematic of plumbing ​

                                                        Figure 7: 3 phase 24V EDF  ​

​Then the heat removed from the generator is simply the mass flow rate x the heat content.

​The generator efficiency is estimated from the relationship between mass flow rate/ heating surface area, the generator has been sized to achieve ~80% efficiency at full power. Since we now have a value for heat removed in BTU/hr the fuel demand is simply calculated from:

approximations within this approach, will be such that equilibrium is unlikely to be exactly achieved, as such temperature and pressure will rise or fall during operation in proportion to the errors within the calculations.
There are various permutations. However, in simplistic terms (for the sake of this article) there are four basic scenarios, temperature to low or too high, and pressure too low or too high.
1. If temperature is low but pressure is good, then there is excess water in the generator, i.e. the water to saturated steam transition is too high up the coils. The remedy is to reduce feed water input.
2. If temperature is too high but pressure is good, then there is excessive heat and insufficient water. The remedy is to reduce burner input, and increase feed water (a normaliser can be used to control excess temperature. However, additional feed water must also be added at the bottom end to remedy the situation.)
3. If pressure is low but temperature is good, then again there is insufficient water in the coils. The remedy is to increase feed water and heat together.
4. If pressure is high but temperature is good, then there is too much water in the coils. However, both water and fuel must be reduced together.
 
Formula for temperature and pressure correction factors have been created. These are used to bias the fuel demand and water delivery. Therefore, the initial fuel and water demand, which was calculated from heat/ steam being removed, is either increased or reduced according to the conditions in the steam generator relative to the operating set point.

Clearly if either the boiler efficiency or the steam usage is significantly out, then the system will find equilibrium away from the desired set point. Therefore, trim pots are provided on the control interface to allow the boiler efficiency and or the leakage factor to be adjusted to bring the convergence in line with the operating point. An error in efficiency would result in issues related to heat input, errors in steam usage would result in issues with feed water rate.

If steam usage was underestimated, then we will see an increase in temperature with a falling pressure. This is because there is a reducing water level in the generator. Therefore, the leakage factor must be increased. Should pressure be high and temperature low, then steam usage is overestimated and leakage factor should be reduced. Note the reduction in calculated steam removal has the effect of reducing both feed water and heat input proportionately. Therefore, pressure should reduce. However, temperature will be maintained via the reduced water level within the coils despite the proportional reduction in heat input.

Should boiler efficiency be underestimated then we will see over temperature and over pressure. In this case boiler efficiency factor should be increased, this will reduce the fuel delivery. Should temperature and pressure be low, then boiler efficiency has been over estimated and the boiler efficiency factor should be reduced, thus increasing fuel delivery.
So leakage factor increases or decreases feed water and fuel in proportion
Boiler efficiency either increases or decreases burner input alone. ​

​There are three control modes:
Minimum Burn:
At its lowest fuel pressure setting and with the return line fully open, the burner delivers approx. 1.2gal/hr. In this setting the burner cuts in and out at the operational point for pressure and or temperature, the fan continues to run at low speed to prevent a damaging heat soak, which could otherwise occur. The system stays in minimum burn, whilst demand is less than the minimum burn value (1.2gal/hr). Speeds in excess of 30mph will be required before demand exceeds minimum burn.

PRV control:
Once demand for fuel exceeds minimum burn mode, the system automatically switches to PRV control. Now the system varies the PRV position in the nozzle return line to match fuel delivery to demand. As the demand increases, the PRV progressively closes. The return flow meter is used to calculate fuel delivery. However, as the PRV approaches fully closed, the low flow starts to make the flow meter readings unreliable. As such, an algorithm is used to calculate return flow from PRV position at the final stages of closure. Unlike in minimum burn mode, the system modulates around the desired operating point. As previously described the fuel demand and water is modified based on generator conditions. Therefore, if pressure exceeds the operating point, this results in a reduction in fuel and feed water. The reduction progressively increases the further from the operating point the system gets. If this results in reducing steam temperature, then the feed water will be cut back further still. If the steam temperature actually rises due to the reduced feed water, then feed water is increased. Should the control system fail to keep the pressure and or temperature within the upper operating band around the set point, the system will drop into minimum burn mode. Once conditions return to within the operating band, the burner will ramp back up to the required demand again. Should, however, even minimum burn mode not prevent further pressure and or temperature rise, then upon reaching the safety limit the burner and feed water will fully shut down.

Speed Control mode:
Once fuel demand exceeds PRV mode, the return line valve and PRV are fully shut. Now fuel delivery is varied by increasing fuel pump speed. This has the effect of increasing fuel pressure in the fuel rail. Fuel pressure can be varied from 100 – 300 psi and at 300psi this gives 9.9g/hr. Modulation (including shut down) is as described for PRV control mode.

Feed Water control
The water pump is driven from the output shaft of the engine via the free-wheeling hub Therefore, even when coasting, the water pump is driven via the back wheel. However, no feed water is available when stationary. The outlet from the feed water pump has a recirculation line back to the inlet port of the pump via a solenoid valve. When this valve is open, no water is delivered to the generator. It is a simple task for the PLC to modulate the solenoid valve to control the proportion of water delivered. In addition, an offtake from the pump delivery goes to a second solenoid valve and from thence to the normaliser. Should the stem outlet temperature be too high, then the normaliser solenoid valve opens and allows cold water to injected to the top coils of the generator. The feed water entering at the bottom has an increased pressure compared to the steam leaving the generator at the top, due to the pressure drop related to velocity within the generator. As such the pressure delta increases with demand and so the fixed orifice in the normaliser nozzle has been sized to deliver approximately 10% of the feed water rate for any given speed and demand. The normalizer nozzle should consist of an orifice followed by approximately 3 inches of shroud within the steam main. This shroud is heated by the superheated steam. The water spray entering the shroud (via the orifice) is heated and turned to wet steam within the shroud before it mixes with the main steam flow, ensuring good mixing and efficient heat transfer between the gaseous fluids. If water is merely injected straight into the super heater coil, then the water droplets bead like water dropped on a hot plate and the effectiveness of the system is much reduced, (remember at full demand the velocity in the superheater is >75ft/sec), with a mix of wet and superheated steam leaving the generator outlet. The water pump is fixed displacement, thus the feed water rate is calculated from engine rpm and validated by a flow meter within the suction line to the pump. A non-return valve on the pump inlet ensures that when in bypass mode, the internal pressure in the pump is largely maintained and not just lost back to the water tank. An electric boost pump ensures that pump inlet conditions are maintained to prevent cavitation.

Part two will follow after further updates to design and testing.

Pat Farrell explaining how he balanced his 30 HP Stanley steam car engine. 
reprinted from SACA  FORUM.
Re: Balancing an old steam engine 
I added additional counterweights to our 30 HP Stanley engines' crankshafts. Before the added weights, the Stanley engines started bouncing at about 47 M.P.H. Just minutes ago, I just blew down our 1911 Stanley model 85 7 passenger 30 HP touring after a 25 mile demonstration ride. I was giving some of my classmates (class of 1963) a thrilling ride at 70 MPH and there wasn't a bit of bounce anywhere coming from the engine. Smooth as silk! The Stanley still had more throttle left to use. The 30 HP Stanley car is unbelievable! The added counter weights made it just that much better.

 The counterweights that I added to our 30 HP Stanley engine were almost the same size as the counterweights that the engine throws already had on it. I do not know their exact weight. The were crafted by REMPCO in Cadillac, Michigan by Gilbert "Red" Hall. Phone number 1-800-736-0108 If one has a 30 HP Stanley and he is not using these counterweights as of yet, then they have not experienced the true smoother higher speed potential of their Stanley. Our model 85 is geared 50 to 60. Almost one to one. The tires are 36" in over all diameter. At 70 MPH, the Stanley engine is smoothly just loafing along.
Our 1911 Stanley model 85 7 passenger, 30 HP is old technology that they already had by 1911 and about 50 years of research in it to make it the most advanced form of transportation of its time. By updating in the proper areas, like engine balancing, modern metals, better brake lining, safety glass, better tires, so on and so on, we presently have one of the most advance forms of steam transportation in this modern era. Our model 85 has never disappointed me in its performance or reliability. I drive with modern traffic. Its handling is on par with many modern vehicles. My only short comings are that about every 50 miles I have to look for water, and I have to be cautious with its two wheel brakes. 
My pride is that this Stanley is that it is all scratch built with scrounged up parts by me. I tried to make it historically accurate too. 
People who scratch build their modern steam cars are unfortunate in that they often are trying to re invent the wheel again by using previous failed inventions. If the modern steam car builders would comb the back issues of our Steam Car Bulletins, they could learn a lot from the failed past ideas used with steam cars. Oh yes. I have had a lot or reworked parts through the years too. Perfection eventually arrives. 

Image above is a conversion done in England by Basil Craske similar to Pats idea.

​

1.  How many of you have heard of the Ashton Valve Company?
2. How many of you remember seeing an Ashton Gauge or Valve in the field?
3. How many of you have experience with a boiler?

Two years ago I knew next to nothing about the Ashton Valve Company. I remember my father mentioning the name a few times over the years but he knew little about the company outside of a few family names and the fact that there was a building in Cambridge with “Ashton Valve “carved on it.
I was sitting with my mom and dad talking about my mom’s side of the family and what I had recently learned about them. My Dad said, “Why don’t you try and find some information on the Ashton side of the family?”. “Sure”, I replied. I immediately found Ashton Valve online. Most of the references were links to Ashton items that had sold on Ebay.
It wasn’t long before I had photographs of long deceased relatives that my Dad had never seen before. My scrapbooks soon filled with hundreds of old advertisements and articles I had downloaded from turn of the century steam trade journals. The book shelves at my house started to fill with old gauges and other artifacts related to Ashton Valve purchased on Ebay. One of my friends who happens to be a steam enthusiast suggested I visit the annual “Steam Up” at the New England Wireless and Steam Museum in Rhode Island. I contacted the people running the museum and they loved my idea of having an Ashton Valve exhibit at the event. I honestly had my doubts that anyone would care about a largely forgotten company. I was wrong. Quite a few people asked me questions about the company and told their stories about working with steam. I had purchased a CD online of old Ashton Valve catalogs and the man who sold it to me surprised me by showing up at the show and giving me a safety valve from 1874, and a stock certificated dated 1877.
That day in Rhode Island was very inspirational to me. Not long after that, the same friend who told me about the “Steam Up” mentioned a museum he had visited some years back that had sponsored a steam show he thought was excellent, and suggested I check out the museum sometime. It was the Charles River Museum of Industry and Innovation. A few emails and a couple of visits later and here I am, talking about Ashton Valve and looking forward to the exhibit that will be here sometime this year.

It’s contribution to the overall sales is one of diminishing proportions, another example to be anticipated when a company does not assume a responsible role in the design development and manufacture of competitive products for an ever-changing market.” Unfortunately for Ashton Valve, the company never ventured far from the production of their steam related products.
Ashton Valve merged with the Crosby Steam and Gage Company in 1948, sold the building to Nicholson & Co., an industrial adhesive manufacturer, and moved with Crosby into the old Winter Brothers Tap and Die building in Wrentham, MA. The Wrentham building was torn down in 2012, Crosby was purchased by Tyco and still operates out of a facility in Mansfield, MA.
​ https://www.smokstak.com/forum/attachment.php?attachmentid=338041&d=1561734973
Rick Ashton

                                                              Engine Block 
Unscrew elbows from the engine block and remove copper covering. It will expose oil soaked insulation which needs scraping off.

                                                                  Valve Cover
The valve cover has a hole in it for the drip valve. The pipe used to help take the unit out of the car can be used as a fulcrum to wrench the cover off. It should be so oil soaked that it should screw off easily after you get it loose. One of the ears is missing which led me to modify a pipe to use against the remaining ear. It worked fine. When reassembling, the cover has packing in the top and needs removed and replaced with a strip of 1/8” packing. The cover was replaced with a new one Don Hoke had in stock in 2019.

​The valve surface should be smooth and not curved. Steam pushes the valve against the block and creates a seal.

                                                                      Gearbox 

The oil level in the gearbox should be about 1 1/2 “ from the bottom while in the car.
Bearings and sliding/turning surfaces should have minimal of .001-.003" clearance. It is easy to over-tighten bolts, and looser is preferred over too tight, especially with cam and crank rods. Make copper shims to space everything appropriately.

                                                                   Crank gear 
The crank gear is significant as it has a larger gear welded on top of a smaller gear. This was a fairly common practice as road conditions improved over the years allowing higher speeds and welding an existing gear over the original one was a low cost option to replacing the entire crankshaft. 

                                                                 Rear Axle
Differential
Removing the rear wheel requires a special puller that fits in the hub. The complete axle with wheels weighs 394lbs.
The main item to keep in mind is that play between the planetary gear and the spider needs to be as close to 0.001" as possible with less being more desirable than more. In the case of this car, there was so much play that the gears barely connected, resulting in catastrophic failure of all three spider gears as well as damage to the case.

                                                                        Axles
Moving the planetary gears in ended up causing the axle to shift inwards on the rear end and the clearance between the wheel and rear-end disappeared causing the wheel to lock up on the driver's side and rub on the passenger side. In 2018, I disassembled the rear end again. Jim Wright built up the steel and I machined the end up a quarter inch further out. The Wheel is a press-fit with the nut torqued to around 200 ft/lbs on the 5deg cone splitting the 1 1/4" and 1 3/8" section. The end result is that the wheel fits better but there is no longer space for a washer under the nut on the end of the axle ,and in the case of one of the axles, a 1/16" cotter-pin hole needed to be drilled at the end of the shaft.

​The end result pushed the wheel out about 1/4" inch. The collar was turned and polished to accept the new oil seal from McMaster: 5154T29 (SKF 17231) for shaft 1.750" and bore 2.250". The collar is put on the axle with some gasket material. One axle has a middle diameter of 1.325" whereas the other is wider at 1.360". I did not change it. The outer bearing is a sealed unit and should not need any care.

​Before and After (2018)

                                                         Oil Pump Adjustment

​The absolute longest would be a point where it almost reaches the intake port. Very little oil will be pumped with this setting. 

The rod is adjusted by moving the nuts in or out and then tightening them against each other.

To set the limit of the primary piston, place the metering rod towards the end of the hole, push in the primary piston until the two meet and mark the spot with a marker.

​Once back in the car, screw the piston into the mount until the most forward motion reaches the mark.

Brass Cleaning 
As There is a lot of Brass/ Copper on this car, polishing it will take a fair bit of time.
Polishing very tarnished Brass/Copper is best done with a soft cloth polishing machine and red rouge for a rough clean. White rouge on a finishing cloth seems to leave a build-up and the finish does not improve.
The likeliness for a Polishing machines to grab and throw parts to the ground at great velocity is directly proportional to their fragility and replacement cost.
Polish 
The best way I have found to clean by hand is through the Wizards product line. A buddy introduced me to Wizards Metal Polish, which is an impregnated piece of cotton. It worked well but not nearly as well as when I added Wizards Metal Renew to it.
Other products tried were :
Mothers Billet Cleaner (Good)
Brasso (Poor)
Most products sold at flea markets and products like Oxy-clean have found their way into the trash bin.
Saving the shine 
As of 2019, I have not found a suitable product that works.
I stayed with Wizards Metal sealant as it came in a package with the other Wizards Products. It seems to help a little. Eastwood has a product in a spray can that almost looks like a clear coat. Neither really kept the Brass from turning yellow and the Eastwood product required the buffing wheel to remove it.

                                                                            Paint
By the end of 2018 the paint started coming off sections of the car in sheets. On other parts like the hood and the front, there was a lot of burn damage. For a short time, I considered stripping the car down completely and have it repainted professionally, but veered away from that as the vehicle is getting a lot of use and the time it would take to do so would probably turn into a full restoration: as well as the slim chance of a fine pint job remaining that way for any length of time lead me to search for an alternative Gloss paint that would be durable. I eventually settled on Sherwin Williams Industrial Enamel. My Neighbour Arthur helped with prep and paint of the large flat section, as I did not want brush marks. The hood is a combination of prep styles as well as some brushing and some spraying.
The paint on the side was in very poor shape, and was not sticking to the wood at all. The passenger side seemed to be less affected by it.

​Early attempts included trials with and without primer as well as brushing VS spraying. Arthur was of great help here.

Eventually, we settled on a sand-able primer and spraying the enamel after making sure the primer was smooth. The gloss is not nearly what I would expect from automotive paint but I will try wet sanding it next winter. It takes a long time to dry.

​The dark blue stripe with the 1/8" light blue stripe within was achieved by putting a 1/8" piece of masking tape and paint the dark blue over it. I found that this paint came out very well if the masking tape was removed immediately after painting, as it allowed the paint to flow. Multiple coats of enamel tend to cause it to crinkle, probably since it takes so long for this paint to dry.

                                                                     Resources
Steam Oil: Siphon-a-min 1000     Brennan oil 970-247-3054
Machining of complex parts:  Rempco 231-775-0108
Boiler repair: Don Bourdon (802) 457-3787

​Tank Sealer: CHEMSEAL B2 TANK SEALANT CS3204
Brass:    Wizards 3004.2982 11011 Metal Polish.
            Wizards Metal Renew_Silver Polish, Aluminum Car Care, Motorcycle Copper Cleaner.
            Wizards 11021 Power Seal Metal Sealant.

                                                                    Steam Tables

                                                                Fuel Filters
​​I installed 2 fuel filters/water separators in 2017 after noticing how much filth got into the system. I feel the pumps, Fuel automatic, Check valve and steam automatic will be well served with clean fuel. In addition to that, the 55gallon drum holding the Jet-A seemed to not get rid of water after rain, so I had some worries there. Parker industries introduced the RACOR SNAPP 2 micron water diesel fuel separator. It is a one piece unit of which the insert is easily replaced. Water settles in a transparent bowl and is easily drained. The valve is the main shutoff at the fuel tank. Fuel was unable to pass through anymore.

​For the Hexane system, I installed a RACOR 110A 10 micron fuel filter. It can be pressurized and it was plumbed in with copper tubing. It should perform to 35 psi

                                                            Steam automatic

                                                          Exhaust System.. 
The initial system had big gaps from where the engine pushed it out before the engine steam intake was shortened. Welded the system closed to where the steam discharge pipe came in. I’m puzzled with the pipe. It is simply soldered together and seems to hold fine

                                                             Tailpipe. 

I had soldered the tailpipe in the past/ it did not hold so I ended up welding it (TIG with High Silicon Brass rod from mcmaster)

                                                              Pilot Light. 
Bottom unit was a Propane setup. I converted it back to Hexane. Pilot body by Don Bourdon

                                                Sight glass (Gauge glass). 
An extremely useful study source is the klinger product catalogue at http://www.klinger.kfc.at/index.php/en/component/jdownloads/send/3-product-catalogues/19-product-catalogue-gauge-glasses
The unit on the steamer is as size 1B reflex gauge and can be found at mc master at https://www.mcmaster.com/#gauge-glass/=1cljafz

              The following came from Pat Ferrell in Washington state after a facebook inquiry:
​
I do use the McMaster reflex gauge gaskets, as they are a good value and they fit well. I use an antisieze on my gauge's bolt threads and I torque them starting from the center bolts like torquing a car's head gasket. Tighten a little at a time. I torque them when the gauge is cold, and then I re-torque them after the gauge has been fired and has cooled down again. I re-torque them every spring as an annual service. How many pounds? I torque them firmly, as I don't use a torque wrench, I don't want to guess the torque. In the 33 years of steaming, I have never had the quality of a reflex gauge glass degrade from use. Through the years I have lost about 3 gauge glass gaskets while on the road. I always carry spares with me.

 
​                                                 Burner (Baker Style. Made by Blazik)
Remove and install: the burner will not come out with the car at its normal level. Lift the body and install two pieces of 2x4 to keep it up. A long enough piece of plywood held up in two spots will help finesse the thing out of the car but is indispensable to tweak it back in
Disconnect the steering bar to give more room

                                                        Burner make-up. 
The burner sits in an insulated pan. The easiest way to take the burner assembly out of the pan and put it back in is by doing it upside down. The gas diffuser is simple but the pipes are above the pan. If the pilot goes out and raw kerosene is blown in, there is no way for it to leave other than by burning it off.
I used muffler repair dope to seal the halves together and furnace cement to seal up everything else
Use masking tape to cover the slits and protect them from cement. Use insulation between pilot and burner
“Imperial hi-temp stove and furnace cement, Grey” seems easy to work with
Everything is screwed together with #10 button heads

                                                                 Hubcap. 
  lost a hubcap in Iowa, decided to make a new one from ultra machinable brass from mc master

                                                                Headlights. 
 
 made headlights operational with a small Acetylene tank and regulator from Trico