# The Design, Construction, and Use of a SMALL PRONY BRAKE

#### Content Tools

11155 Stout Road, Amanda, Ohio 43102

When I decided to investigate the possibility of constructing a Prony brake I don't believe that 1 had ever seen one. And I know that I had never see one in use. My only knowledge of them came from textbooks. I approached the project by reading every old text that I could find that had any reference to the Prony brake or dynamometer. I needed to find the answers to such questions as:

How big should the brake drum be to measure 10 horsepower?

How big will the scale need to be to measure that horsepower?

What is the best material to use for the brake lining?

How long should the arm be that goes from the brake drum to the scales?

The following is a brief summary of some of the information that I collected.

### Some History of the Prony Brake

The concept of rating engines, water wheels and windmills in terms of how many horses they were equivalent to, dates back to the early 1700s. It was not until the late 1700s that anyone made any real effort to determine just how many foot-pounds per minute a horse was capable. This determination was made by James Watt, the father of the modern steam engine. Watt was selling steam pumping engines and needed a reliable figure to use when he was attempting to sell a client on the idea of replacing horses with one of his steam engines. However, Watt was also concerned that his engines develop a reputation of being more powerful than any of his competitor's engines of the same horsepower rating. To do this he simply inflated the number of foot pounds per minute in a horsepower to 33,000 from the approximately 25-27,000 that his research showed. Thus a five horsepower Watt engine would do significantly more work than five horses. This inflation in the numbers created problems for some users of gas and steam engines for over 125 years as the builders of traction engines continued the practice of under-rating their engines. Because Watt was selling pumping engines he was able to calculate the horsepower using the number of gallons pumped, the height that the water was raised and the time that was required for a given volume. However, once he started building 'rotative' engines (engines with a crank and flywheel) he did not have any reliable method of calculating the output horsepower. Around 1800 he invented the steam engine indicator and he may have used it to calculate the input horsepower of his engines. But because of the relative low efficiency of the engines and the primitive construction of the indicator, it probably could have provided only a rough estimate of the power delivered to a rotating load.

The problem of how to measure the horsepower of a rotating shaft was solved by Gaspard de Prony in France in 1826 when he invented the first friction brake. This device came to be known as the Prony brake. The following sketch, (redrawn from an 1868 book) shows the simplicity of de Prony's invention.

### How the Prony Brake Measures Horsepower

In use, the stationary lever 'D' is clamped around a rotating shaft 'A' and the two bolts above 'A' are tightened until the engine is working up to full load. Weights 'B' are then added to the scale pan until the lever 'D' drops slightly away from the upper stop 'C.' The only things that needed to be known to calculate the horsepower were the length of the lever 'D,' the weight of 'D' at its right hand end, the additional amount of weight 'B' added to the scale pan and the speed of the shaft 'A' in RPM.

Foot Pounds per Minute = Pi x 2 x Length of D x RPM x Weight

Because Watt's figure of 33,000 foot pounds per minute per horsepower has survived through both the 19th and 20th centuries:

Horsepower = Pi x 2 x Length of D x RPM x Weight/33,000

For an excellent discussion of the concept of horsepower see the late Amos Rixman's article in the April-May 1989 issue of Gas Engine Magazine.

### The Prony Brake Constant

Because all calculations had to be made either with pencil and paper or with a slide rule, many Prony brakes were designed to simplify these calculations. Because the factors Pi, 2, Length of D, and 33,000 were constant for any specific brake, they could be combined into one factor. This factor was known as a Prony brake constant. If the length of 'D' was made to be 63 inches, this factor would be exactly 1,000. Using this factor, the above question for horsepower become simply

Horsepower = Weight x RPM/1,000

On the smaller Prony brakes, up to about 40 HP, the length of the arm was sometimes made 31? inches yielding a Prony brake constant of 2,000. This is the length that I used when I built my Prony brake.

### Designing the Brake Drum

The most difficult question that I had to tackle was; 'What size of brake drum is required to measure 10 horsepower?' Answers were not easy to find and when I did find them there was not much agreement from one authority to another. Halsey, in 1916, stated that 14-4 square inches of drum surface was required for each horsepower. Moyer in 1934 felt that only five square inches was needed. The Standard Handbook for Electrical Engineers in 1915 stated that, 'About 100 square inches of rubbing surface of brake should be allowed with air cooling, or about 50 square inches with water cooling per horsepower.' These conflicting figures were obviously of no help. Mr. Flathers, in 1892, commented that the size of the drum is important because if the surface is too small the operation of the brake will be irregular and if it is too large, considerable weight may be added to the scale without materially altering the position of the lever arm. I eventually compiled data on 18 Prony brakes that were described in sufficient detail in the old texts. When I listed these brakes in the order of their rated horsepower (See Table 1) I discovered that the number of square inches per horsepower decreased dramatically as the size of the brakes increased. For example:

A two horsepower brake was built with 113.1 square inches per horsepower.

Seven brakes ranging from 19 to 40 horsepower had an average of 58.3 square inches.

Seven brakes ranging from 125 to 183 horsepower had an average of 21.2 square inches.

A 250 horsepower brake was designed with 14.5 square inches.

A 475 horsepower brake had 13.9 square inches.

A 540 horsepower brake had only 8.4 square inches per horsepower.

A closer look at the data, however, disclosed that conclusions drawn from this information could be very misleading. One example, Prony brakes with areas from 1885 to 1979 square inches of drum surface were listed as having measured horsepowers ranging from 33 to 140. Some of the capacities cited are the capacities of the engines, not of the Prony brakes. Some of the texts proposed a 'K' factor that took the velocity of the surface of the drum into consideration. I found these to be confusing but I included them in Table 1.

When I sorted through the pulleys that I had available to make a brake drum I found that the best pulley was 12 inches in diameter by 5/8 inches wide. This yielded an area of 211 square inches. Based on the above information I have concluded that this pulley as a brake drum, (28 to 42 square inches per HP) my Prony brake will probably be limited to 5 to 7? horsepower on short runs and even less on long runs. This fell shy of the 10 horsepower that I had been shooting for but seems to be a very reasonable size for most hit and miss engines. I have not yet really put it to the test to determine just how much horsepower it can absorb. The brake drum can be seen best in Figure 6.

It appears that the capacity of a Prony brake is determined by the amount of horsepower it will absorb without the brake blocks breaking into flames. On this basis, capacity for short runs would be much greater than for prolonged runs.

### Selecting the Material for the Brake Lining

#### Specification for 18 Prony Brakes that Provided the Data for My Design

 HP RPM FACE Inches DIA Feet Length of arm DESIGN OF BRAKE K Square inches Sq. per In. HP Avg. Sq. In./HP Velocity at drum Ft./Min. Avg. Velocity 2 2 200 3 2 Flathers 226.2 113.1 113.1 1256 19 148.5 7 5 33 McLaren, comp 858 1319.5 69.4 2331 20 146 7 5 32 McClaren, water-cooled & comp. 802 1319.5 66.0 2292 21 150 7 5 33 Royal A. Soc, Comp 785 1319.5 62.8 2355 33 150 10.5 5 32 Garrett, water cooled and comp. 749 1979.2 60.0 2355 38 378 13 4 Flathers 1960.4 51.6 4748 40 180 10.5 5 32 Garrett, water cooled and comp. 741 1979.2 49.5 2826 40 30.1 322 13 4 28 Westinghouse, water-cooled 847 1960.4 49.0 58.3 4044 2993 125 290 24 4 63 Westinghouse, water-cooled 465 3619.1 29.0 3642 125 290 13 4 28 Westinghouse, water-cooled 847 1960.4 15.7 3642 140 350 Halsey (1903) 14.4 0 140 250 20 2.5 Kent 285 1885.0 13.5 1963 150 150 10 9 Schoenheyder, water-cooled 282 3392.9 22.6 4239 150 60 24 5.5 Francis 4976.3 33.2 1036 183 144.7 249 24 4 Flathers 3619.1 19.8 21.2 3127 2942 250 250 24 4 63 Westinghouse, water-cooled 465 3619.1 14.5 14.5 3140 3140 475 362.5 76.2 24 7 191 Webber, water-cooled 84.7 6333.5 13.3 13.9 1675 1675 540 100 24 5 126 Getely & Kletsch, water-cooled 209 4523.9 8.4 8.4 1570 1570

Like many other characteristics of the Prony brake, I found considerable disagreement as to the best material to use for the lining on the brake band. Kent's Mechanical Engineer's Pocket-book, 1910, 'Soft woods such as bass, plane-tree, beech, poplar, or maple, are all preferred to the harder woods for brake blocks. This recommendation appears to be consistent with the most common designs. However, at least one Westinghouse brake used hardwood blocks of either oak or hickory. Some other materials that were mentioned included rope (which was widely used for a variation of the Prony brake), babbitt, cork, woven belting, asbestos, leather and even linen.

I followed the advice from Kent's Mechanical Engineer's Pocketbook and cut the brake blocks for my Prony brake from soft pine, 2 x 4s.

### Cooling the Brake

For prolonged testing, most of the texts recommended that the brake be designed with inverted flanges so that the inside of the brake drum can be flooded with water that is held against the drum by centrifugal force as the drum rotates. To assure that a constant amount of water is contained within the flanges a continuous stream of water is introduced. To prevent the water from overfilling the flanges, a scoop was often installed that removed a steady stream of hot water as the cooler water was added. This, like most aspects of the design of the Prony brake, failed to receive unanimous agreement from all of the authorities. Mr. Ludy, in 1912, felt that a scoop was not necessary because the water would never reach the boiling point. However, a photo in a book from 1907 shows a Prony brake completely engulfed in steam from the heat generated during a test on an electric motor. The authors of that book, Mr. Swensen and Mr. Frankenfield, claim that the best results are obtained when the water is added just fast enough to make up for the evaporation from the boiling water. Others reported that when a brake was run near its capacity it was necessary to have a man standing by with a garden hose because the drum would get so hot that the wooden brake blocks would break into fire.

The author of one text commented that if water was not supplied to the brake drum, the heat could build up to where the expansion of the rim would be sufficient to break the cast iron spokes.

I incorporated the water circulation feature into my Prony brake by installing a small antique American-Marsh centrifugal pump that is driven by a flat belt from the main shaft of the brake. To assure that the pump does not consume power that is not registered on the scale, I supported the pump from the brake band. By doing this, the torque that is required to drive the pump is automatically added to the torque that is required to overcome the friction of the brake. I also included a scoop to remove a steady stream of warm water from the drum. This water is collected in a small reservoir and recirculated through the pump jack back into the brake drum. If I ever run the brake very hard I may decide to add some provision for evaporative cooling. The pump can be seen in Figure 6 and the reservoir in Figure 5.

### Designing the Lever Arm Between the Drum & the Scales

In the 1870s a professor named John E. Sweet made one of the few improvement on de Prony's design. Professor Sweet developed a design for the lever arm that consisted of two pieces, as opposed to the single piece that de Prony used. This made for a much lighter arm, which reduced the weight on the drum and bearings, and it also provided a more balanced force on the brake band. Examples of both as brake of the original design and one of 'Sweet's Pattern' are shown in Figure 2.

When I built my Prony brake I elected to use Professor Sweet's design. The arm can be seen in Figure 5.

### Measuring the Speed of the Drum

Apparently the early tachometers were not very reliable or at least the authors of the texts did not have much confidence in them. One author wrote in 1919:

'Speed in revolutions per minute should be invariably taken from positively driven counters which engage at the beginning of the run and disengage at the end. The difference between the two readings, divided by the duration of run in minutes, gives the true average speed. Tachometers, even though carefully calibrated, are not sufficiently reliable for RPM readings. In connection with the speed counters mentioned, however, the tachometer may be used as an appropriate check on average speed, also as an indicator of variation in speed before or during the run.'

Even though I own a very nice digital tachometer, I chose to mount an old mechanical machine counter and a stopwatch on my Prony brake to measure the speed of the drum during a test. The counter is actuated by a lever that is in contact with an eccentric on the outboard end of the drum shaft. The counter and stop-watch can be seen in Figure 5.

If I ever find a reasonably priced antique belt-driven tachometer, I may eventually add it to indicate variations in speed when checking the operation of engine governors. This check is performed by rather quickly adding and removing a load and observing how much the speed of the engine varies. This would be very difficult (probably impossible) to do with a mechanical counter.

### Determining the Size of the Scales

To determine the size of scales that would be required I needed to know how fast the drum would be turning and how much horsepower I would be measuring. This was a trial and error process. I first listed the speeds of my gas engines and the diameter of the pulleys on each. I then looked at what sizes of drive pulleys I had that I could use on the brake shaft. Then used the equation that I mentioned earlier:

Horsepower = Weight x RPM/2,000

The force on the scale is:

Weight = Horsepower x 2.000/RPM

For example, my 1? HP Fairbanks-Morse engine that has a pulley 5 inches in diameter would turn the brake, with a 14-inch pulley, 268 RPM

When I put 1 ? HP and 268 RPM into the equation, I found that the force on the scale would be equivalent to a weight of 11.2 pounds.

Then I looked at what would happen if I was to put my 8 HP United engine with its 18-inch pulley on the brake, I found that it would turn the brake 579 RPM.

When I put 8 HP and 579 RPM into the equation I found that the force on the scale would be equivalent to a weight of 27.7 pounds.

I thus determined that I would need a scale that could accurately measure anything from 10 pounds to at least 30 pounds. I had originally assumed that I would use an old platform-type granary scale but this calculation showed me that it would be better to use an old tabletop Fairbanks scale that is considerably smaller.

### Putting All of the Pieces Together

One of the challenges in building my Prony brake was to find a way to construct it from the pieces that I could find in and around my barn. After selecting the scales, the brake drum, was the biggest challenge. I wanted a drum with the inverted flanges so I could apply water to the inside of the drum but I certainly did not want to go to the expense of having a foundry make a pattern and cast a special drum to my specifications. The first thing that I did was to decide on the number of square inches of surface that I would need and then I looked for the cast iron pulley that came closest to that size. Then 1 cut two steel rings from a piece of 3/8' steel plate and machined them to fit against the two sides of the pulley. I also took a light machine cut off the sides of the pulley so that I would have a good fit, applied a little gasket material from a tube and clamped the two plates to the pulley with long bolts running from one side to the other. I left about a quarter of an inch flange on the outside of the pulley to assure that the brake band would not ride off the pulley.

The wheels were easy, I didn't have many to choose from so I simply used what I had. It would have been nice if I could have had a matched set of four, but I didn't. At least they match from right to left.

I built the engine mount at the rear of the frame so that it is universal in that I can match almost any arrangement of engine mounting holes. The problem with this arrangement is there is no easy way to get engines on and off from the frame without some kind of hoist. This didn't turn out to be a big problem because I found that I could run the belt the other way and hook up to an engine that is setting on the ground or on its own trucks.

One of my goals in designing and building this brake was to make it look as much like a vintage piece of equipment as possible. To do this I refrained from using any welding and I did not make any effort to remove rust from any part unless it would interfere with the operation of the machine. I finished the wood parts by merely coating them with oil.

One thing that is important in building a Prony brake is to be certain to include a back-stop, especially if you expect to be testing any steam engines that could be started in reverse. Someone could get hurt or the equipment damaged if the lever

arm was to accidentally fly over backwards because someone forgot to twist the belt or accidentally started a steam engine in the wrong direction. The back-stop can be seen between the scales and the support for the counter in Figure 5.

Prior to starting construction of the Prony brake I made a drawing to guide me as I put it together. Needless to say, the drawing involved along with the construction as I selected parts and as I solved problems that arose. Figure 3 is the end result. Figure 4 shows how the Prony brake looks today.

### Lubrication of the Brake

Like many previous facets of the Prony brake, there were a variety of recommendations regarding the proper lubrication of a brake. Recommendations ranged from water, to tallow, including soap-suds, oil, and heavy grease along the way.

The drum of my brake is lubricated with a heavy grease that is applied via a large grease cup that is screwed into one of the wood brake blocks. This block has a slot cut into the underside of it to distribute the grease across the face of the drum. The grease cup can be seen in Figure 5.