The KY-2000 Stirling Engine Manual. and Experiment Guide by ttn74823


									      The KY-2000 Stirling Engine Manual.
            and Experiment Guide

                             American Stirling Company

The KY-2000 Stirling engine is intended for physics and thermodynamics classes at
either the advanced high school, or university level.

The engine is electrically heated for accurate power input measurements (and to keep
fumes out of the lab) and is water cooled. The engine has built in pressure, temperature,
and angular position sensors to allow a PV diagram to be plotted on a PC. The engine
also includes a data acquisition system that ports ascii data to a P.C. where it is analyzed
in a Microsoft Excel file. An Excel template is supplied. A variable voltage DC power
supply is available as an option if your lab needs one.

Stirling engines are a type of heat engine that makes an exceptionally good teaching tool.
Not only does the Stirling Cycle match the Carnot cycle for efficiency, but real Stirling
engines can be made to run on relatively low temperatures making them quite good
laboratory tools for teaching about heat engines in general.

     General overview of KY-2000 Stirling Engine

Components of KY-2000 Stirling Engine.

A.    A 30 g. spring scale used in measuring torque and power output.

B.    Working sealed cylinder of engine.

C.    Helium or other non-flammable gas ports. These allow investigation of operation
      of this engine with gasses other than air.

D.    Electrical power for the electronics.

E.    Cable to P.C.

F.    Cold water box.

G.    Flywheel

H.    Variable voltage dc input for heaters.

I.    Spinning output shaft

J.      Hole that allows power measurement by raising a weight or applying torque to the
        spinning output shaft.

K.      Pinch clamps (two). Opening these clamps allows helium or some other non-
        flammable gas to be tested in the engine. These clamps must be closed tightly for
        the engine to operate. Leave these clamps open when the engine is not in use.

L.      Drain clamp. Close this clamp tightly before putting water in the cold water box.

The basic procedure for running the engine and taking data under no load conditions is as
follows. If you are planning a short demo run that lasts less than five minutes water is
not required. However maximum speed is only achieved with cold water in box (L).

1.    Plug in the connectors at D & E. Make sure that both clamps K are closed tightly.
     There are two K clamps but only one shows in this picture. These clamps must be
     closed TIGHTLY for the engine to run.

2. Plug the cable from location E on the engine into the P.C.’s parallel printer port.
   Copy the data acquisition software into a new directory called KY_DAQ.

3. Verify that the variable voltage d.c. power supply is turned off then plug in the
   banana leads at H. Make sure to maintain the correct polarity (red to red & black to
   black) of the cable.

4. Set your power supply to the “control voltage” mode then turn it on and set the
   voltage to 12 volts.

Warning: This engine is electrically heated. The resistors
shown at M get very hot!
The plate that the resistors are mounted on also gets very got. Be careful. Don’t touch
this part of the engine during operation or for at least ten minutes after the power has
been disconnected.

4. (cont.) After the power has been turned on it will take three to five minutes for the
engine to get hot enough to run. When you think it is ready to start, try spinning the
output shaft “I” counterclockwise by hand to try to start the engine. If the engine is not
hot enough, no amount of spinning in the world will make it start. When the engine is
hot enough it only takes a modest spin to make it take off and run at several hundred rpm.

This engine spins counterclockwise when you are looking at it as shown on page 2.
Question: Does it hurt the engine to spin it backwards? Not at all. You cannot harm the
engine by spinning it gently backwards. If the engine is manually spun backwards when
it is hot and if it stops on compression it may come to a complete stop, reverse itself, and
then take off and run in the proper direction.

5. Click on the DAQ icon to start the data acquisition process. The first screen you will
   see is shown below.

6. Click okay and proceed to the next screen which is shown below.

This screen should be self explanatory. Click Yes to continue.

7. When you come to the rather cryptic screen show below, click okay.

8. At the following screen you get to do something useful. Click the box that says use
   Pentium CPU clock counter. Change the file name to something meaningful to you.
   Re-set the sample rate to 1000 and set the number of samples to 3000. There is no
   obligation to use a sample rate of 1000 hz and take 3000, but the templates supplied
   are set up to work with that size block of data, so if you use another rate you may get
   unpredictable results.

9. When the engine is running and ready to be tested, press the start logging button to
   record your data. The file will be located in your KY_DAQ directory along with the
   data acquisition software. Remember that when additional data samples are taken it
   is necessary to re-name your acquisition file or the DAQ software will cheerfully
   create and re-write over successive copies of your file.

10. Open your data file in Excel and convert it to Excel format. Next copy the data to
    the sheet labeled “Raw Data” in the supplied Excel file called “KY-
    2000_Template.xls”. Paste your data into cell A1 of the Raw Data sheet.

                              Power Measurement
Power inputs into this engine can be measured by reading volts * amps off your variable
voltage power supply. Channel 5 in the data stream also measures the voltage drop in
millivolts across one of the 1 ohm resistors. Read the bottom of the Torque & Shaft

Power worksheet for an estimate of how much of this power actually goes into the

Shaft output power can be measured one of two ways. Either by raising a weight from
floor level to the bottom of the engine and timing how long this takes, or by suspending a
thread from the spring scale, then wrapping it around the output shaft in such a way that
the spinning output shaft wants to raise the weight. This is a simple method for
measuring power (you have just built a simple break) and is quite easy to use although
not as emotionally satisfying as watching the weight be raised from the floor. The torque
applied to the shaft is the difference between the spring scale reading and the mass of the
weight * the radius of the shaft.

Both methods are described in detail with examples in the Torque & Shaft Power

Here is a sample set of data for the method of raising a weight.

   Power run #1                                   height       average
5-Mar-00                                           from        speed of
 Grams power (w)                       sec      floor (m)    weight (m/s)
   10     0.013551                      5        0.6858        0.13716
   22     0.016563                      9        0.6858         0.0762
   38     0.021456                     12        0.6858        0.05715
   48     0.020327                     16        0.6858       0.0428625
   57     0.018391                     21        0.6858       0.0326571
   70     0.01395                      34        0.6858       0.0201706

                               Lifting Mass Power run #1
   Power in Watts

                                                                   power (w)
                           0     20     40     60       80
                                 Raised mass in g.

Do you want more power? No, not where students fingers are involved. Twenty
milliwatts or even half of that is plenty of power.

It’s important to know that when raising a weight the thread needs to be guided onto the
spinning shaft by hand carefully so that it does not fall off the end. If you accidentally tie
some knots on the shaft, simply cut the thread off and try again.

Notice that when using the spring scale/break method that the thread going to the scale
goes slack after enough turns are made around the shaft. If you are continuing to suspend
some weight on the spring scale it must be subtracted from the hanging mass weight in
order to obtain the correct torque. The diameter of the output shaft is .125 inches or
.3175 cm. To obtain the correct torque it is appropriate to use the radius of the shaft, not
the diameter.


The data collected by the data acquisition device should be copied into cell A1 of the
template. Obviously it would be handy to know what all those numbers mean. The
Headers worksheet contains a detailed listing of what all the numbers are.

                                        The Lab:
1.     Read pages 1-7 of this booklet and do the pre-lab.

2.     Optional: If you have time plug the engine in and make a calibration run with the
       engine not running and valves K (both of them) open. Record the average
       number from channel 4 for use in the optional indicated power portion of the lab.

3.     Decide whether you will lift the weight to measure power or use the break

4.     Whichever method you use, you will want to take about 10 to 20 data points
       depending on how good your data looks.

5.     Read the Torque & Shaft Power worksheet in the supplied template and make
       sure you understand how to measure power.

6.     Close the drain valve “L” and fill the cold water box “F” with a mixture of water
       and ice. Stir the ice mixture after every data sample is taken.

7.     Collect your data and plot it in Microsoft Excel. Does it look reasonable? If your
       engine is producing 20 kW something is wrong. Plot power vs. RPM and torque
       vs. RPM.

8.                              For data runs with the engine speed at 120 rpm (two revs per second) or slower
                                it’s easier to count the revolutions with you eye and a stopwatch than it is to use
                                the computer.

9.                              Optional: A power indicator graph can be plotted by using the Indicated Power
                                worksheet in the template. This is mostly a manual process but results in
                                indicated power graphs like the one below. The area inside the graph is energy
                                and it must be divided by the time necessary for one revolution to get indicated

                                                       Indicated Power

      Relative Pressure in Pa

                                -200 1.0    1.0     1.1     1.1     1.1     1.1
                                  800E-04 900E-04 000E-04 100E-04 200E-04 300E-04
                                                                 Volume in m^3

These graphs are not automated. They must be produced by plotting corrected pressure
and corrected volume around exactly one revolution. Phase angle must be adjusted by
verifying that bottom dead center (BDC) occurs at minimum cosine of revolution.
Constants must be added or subtracted from cell F1 to produce a PV diagram which is
correct and slopes down and to the right. Add constants to cell F1 until the cosine of the
cell at BDC is –1. Then add increments of Pi/2 if necessary to produce a pv diagram that
slopes down and to the right.

10.                             List three sources of experimental error and estimate their significance in this

11.                             Compare Carnot efficiency for this engine with measured efficiency.

12.    Further experimentation with filling the engine with Helium or another non
       flammable gas is possible by opening both valves “K” and purging the new gas
       through the engine. Never connect these ports to any pressure source higher than
       1.5 psi. You will blow something up.

Comments on measuring power with a thread around a spinning shaft:

       By Darryl Phillips

       One HP consists of lifting 33,000 lbs. one foot in one minute, or 550 in a second.
       Or any combination of pounds and distance that yields that product.

       Now consider a weightless string fixed to the rotating shaft, with a weight
       hanging on the end of the string. It is winding up as the engine runs, but we have
       an infinite amount of this really great string! If we know the weight, and measure
       the distance this weight was raised in a known time, we can calculate horsepower.
       No need to know the shaft radius in this case. Also, the number of turns of this
       infinitely thin thread which have been
       wound onto the shaft is not germane.

       Next step is to use a string that is short enough that we can assume it to be
       weightless. Hang a known weight on the bottom and secure the top end of the
       string to a fixed object, so that the string barely touches the rotating shaft. No
       need for a spring scale at all.

       As the engine runs it will attempt slightly to lift the weight, but the coefficient of
       friction will not be sufficient because there is not enough contact between the
       string and shaft. The question of radius versus circumference is obvious in the
       case of zero turns.Circumference is not involved, nor is it involved with multiple
       turns of string.

       Now wrap one turn of string around the shaft. You have increased the friction
       and the weight will ride a tiny bit higher, but probably not enough to slow the
       engine much. The spring scale should also read a smaller number

       Add another turn, and another, logging the speed change of the engine as
       the friction increases. In each case, rotation will raise the weight just
       enough to loosen the turns, reducing the friction. At some point, adding
       another turn or two will produce no change, you have reached the correct
       operating point for this combination of string and radius and shaft

       At this time, power will compute exactly as you stated. And no spring scale
       would be needed at all. Instead, what is needed is a set of known weights. I
       suggest using split lock washers because they are cheap and available and
       uniform. They can be accurately weighed, with much more accuracy than a

       spring scale is capable of. If necessary the split can be spread slightly so they will
       easily slip onto the string.

       Each added weight will result in a reduced engine speed. Curves of power
       and torque can be easily plotted. These curves will be independent of the
       type of string used, teflon string [Glide dental floss should work well] will simply
       require more turns than will string made of high friction rubber.

       Realize that with zero friction between the shaft and string (as
       in the zero turns case above) the scale will show the weight at any shaft
       speed - i.e., you are not measuring engine performance at all. Obviously
       this is because there is insufficient friction. Again, increase the turns
       around the shaft, one by one, until you reach the point where there is no
       further change and the shaft is lifting all the weight. What reading will
       you have on the spring scale at that point? ZERO, exactly what experiments
       show. (Well, you will have the string that exists above the shaft, but that is
       insignificant and is a constant.) So the spring scale is really unnecessary.

       In any of the above cases, the results will only be reliable IF there is
       enough friction between the string and the shaft. That is why I started
       with the absurd example of zero turns, it applies equally to any of the

After the lab is over: Drain the ice water and fish out the ice by hand, being careful to
not get water on the electronics. Wipe up any excess water with a dry cloth. Open both
valves K. Disconnect all cables and return the engine to storage.


The only maintenance that should be necessary is to put an occasional drop of WD-40 or
similar very light machine oil on the brass linear bearing that goes in and out of the

Warning: Never Never Never oil the graphite piston in the glass cylinder. You will ruin
the engine. The graphite piston is self lubricating.

Warning: Never connect any pressure source to the helium ports “C” that is higher than
1.5 psi. In other words never fill the engine with a gas from any source of higher
pressure than a child’s balloon.

Keep the engine in a clean dry environment out of sunshine. Sunshine will fade the
anodizing on the aluminum.

This engine was designed and hand made by Brent H. Van Arsdell with machining help
from Jim West and electronics help from Bob Nuckolls III.

Brent Van Arsdell has an extremely strong interest in keeping it running. For support
please contact him at Look him up on the Internet if
necessary. Home phone is 316-729-1661. Parents 480-985-1384 or 815-725-0170.
Brother (Lance) 480-654-1798

American Stirling Company


To top