Member postings for Turbine Guy

I received the Housing 4 and Cover 4 prints from Shapeways shown in the following photos. I measured a few dimensions on each print and added the as printed dimension in parenthesis below the design dimension shown in the following drawing.

I read through the section on steam engines in Mark’s Standard Handbook for Mechanical Engineers, tenth edition, by Eugene A. Avallone and Theodore Baumeister III. They stated that a diagram factor of between 0.7 and 0.95 could be used to estimate the actual indicated power from the theoretical indicated power. The theoretical indicated power can be estimated relatively easily as shown on the following spreadsheet. The mass flow into the cylinder can also be estimated with the input data shown. I updated the test sheet shown in the 19/11/2020 post to include the latest tests of my steam engines shown in the 22/07/2022 post using the data from the spreadsheets like the one shown below of each of the steam engines and using a diagram factor of 0.8. You can see that the results look very similar to those found using the methods described in the last post.

As I stated in the 22/07/2022 post, estimating the leakage loss and the mechanical efficiency is very time consuming. The following spreadsheet is what I used for estimating these for the last test of the Chiltern steam engine. I estimate the pressure change required to fill the cylinder and empty the cylinder in the time available. I use these pressures to calculate the enthalpy and specific volume of the gas at each step of the cycle. For air, these are calculated based on an isothermal process. For steam, I use the steam tables and Mollier chart from Thermodynamic Properties Of Steam by Keenan and Keys, first edition. I estimate the leakage, compression, and the friction losses for these pressures. This was the way I estimated the data shown in the test sheet in the 19/11/2020 post. I think estimating the leakage and mechanical losses are important to understand the strengths and weaknesses of different types of steam engines. I decided to try a simpler method to estimate these losses which I will explain in the next post.

I noticed that the drawings showing the piston and valve locations I had added to the Chiltern Steam Engine folder were for the modifications I thought would improve the performance. These drawings are shown in the posts starting with the 01/02/2019 post. I deleted these drawings from that folder and added similar drawings which are for the unmodified version that was tested.

I created photo albums for all my steam engines. I hope this makes it easier to see the details of an engine when it is mentioned or shown in a test report. The following photo shows the steam engines I recently ran with the twin cylinder airbrush compressor mentioned in the last post. The Stuart ST is on the left, the Chiltern is in the center, and the MSM Tyne is on the right. I left out the Saito T1 steam engine since the smaller airbrush compressor I had been using had enough energy to make the pressure higher than it was designed for. The test sheet shown below is a simpler version than shown in the 19/11/2020 post. Estimating the leakage loss and the mechanical efficiency is very time consuming. I can add this to the test sheet if this is helpful.

It has been over a year since I added anything to this thread. All my testing since the last post has been with turbines and posted in the Model Turbines thread. If you start with the 14/05/2021 Post and read the Model Turbines thread to the last post, it will bring you up to date on all the changes and testing I have done since I last updated this thread. The following chart summarizes the results of the testing done during this time. I added folders in my photos album showing the pictures, drawings, and test results for each of these turbines.

I bought a twin cylinder airbrush compressor described in the 14/08/2021 Post that is capable of spinning the propellers used on my steam engines to speeds over 1000 rpm. This airbrush compressor has approximately twice the output energy of the one I was previously using. I will update this thread in the next posts adding the test results of running my steam engines on this larger airbrush compressor and adding folders for each of my steam engines.

In order to get better testing of the effects of the side clearance on efficiency, I decided to make a Housing 4 and Cover 4 to be used with Axial Rotor 3 to form Axial Turbine 4. The following drawing shows what I plan to try. Housing 4 and Cover 4 will be printed in aluminum by Shapeways. This is a selective laser melting process that uses a laser to scan and selectively melt metal powder particles, bonding them together and building a part layer by layer. I decided to try the aluminum since it is much more rigid than nylon and costs about half as much as cast bronze. I hope the finish ends up relatively smooth when I machine the printed parts. The GWS EP 2508 propeller used in the tests pushes the rotor toward the cover plate until this gap is closed.  That results in a clearance between the cover and the rotor of 0.003” and between the rotor and the side plate of 0.003”. In all my testing with the dental ball bearings, they have worked best with around 0.004” to 0.006” of clearance between the set screw collar and the outer ball bearing. The drawing will be updated to show the placement of the parts, nozzle size, and final dimensions that give the minimum pressure to turn the propeller at a speed of 28,000 rpm. 

Edited By Turbine Guy on 11/07/2022 19:06:27

The free version of Onshape has worked very well for me. I do all my designs in 3D and most of the methods used by Onshape seem very intuitive. I have not had any issues with it working on the cloud and feel that since it does all its updates internally, this is a big advantage.

Hope it will work as well for you,

Byron

I made the side plate and positioned it as shown in the following photo and drawing. I got the best performance with a gap of 0.006” between the set screw collar and the outer ball bearing. The GWS EP 2508 propeller used in the tests pushes the rotor toward the cover plate until this gap is closed. That results in a clearance between the cover and the rotor of 0.005” and between the rotor and the side plate of 0.009”. In all my testing with the dental ball bearings, they have worked best with around 0.004” to 0.006” of clearance between the set screw collar and the outer ball bearing. With these gaps the minimum pressure required to turn the propeller at 28,000 rpm was 19.5 psig, the same as without the side plate. I tried changing the clearance on both sides, opening or closing the clearance in small increments from 0.002” to 0.012”. but no combination performed better. The side plate removed the open space the air leaving the rotor had available to flow into and forced the air into a much smaller channel. This may have offset any gain in performance the tight clearances on each side of the rotor might have given. Axial Turbine 3A is the version without a shroud or tight clearance on the rotor OD, so the flow escaping the blade ends might have increased with the smaller exhaust space.

I might try getting some of the lubricant that is used in the dental handpieces for the bearings that are relubricated daily and see how that works for the metal ball bearings I am currently using in the drag turbines. Since these lubricants are designed for the very high speed dental turbines, they may not have the high viscosity at the low temperature caused by the expansion of air in the nozzles. Also, since they are designed to last for an entire day of running the dental drills, the lubricants may adhere to the metal longer. I’ll check the availability and cost of these lubricants and order some if the cost is reasonable. This is all I plan to do with the drag turbines for now.

I plan to add a side plate on Axial Turbine 3A next to the inner face of Rotor 5 as shown in the following drawing. The dimensions for the clearances between the rotor and the cover, the rotor and the side plate, and the distance (x) are what I plan to try. These may change after I find the optimum position of the rotor with the side plate added. The intent of adding the side plate is to reduce the rotational (windage) loss by having a close fit on both sides of the rotor based on what I found for the test described in the 10/05/2022 Post.

My theory for why Drag Turbine 4 ran faster with the exhaust blocked than with the exhaust open is as follows. The gap between rotor faces and the housing or cover is very small (less than 0.002&rdquo. With the exhaust open, the force tilting the rotor is large enough to open the gap around the inlet and the leakage becomes very large. With the exhaust port blocked, the air must flow around the rotor, so the pressure is about equal on both sides which reduces the loads on the ball bearings and keeps the rotor from tilting enough to cause contact with the cover or housing. Eliminating any contact and reducing the bearing loads appears to add more torque than forcing the air to flow out the small vent hole reduces the torque. I tried running Drag Turbine 6B and had the same results. In the last test of Drag Turbine 6B, the rotor stopped spinning and finally the turbine would not run again with exhaust open or closed. Apparently, the forces on the dental ball bearing are too large with the drag turbines. I added a section view of Drag Turbine 6B showing the location of the vent hole.

I pressed the sleeve out of the drag turbine housing and tried the larger all metal ball bearings I had been using. With the rotors in the same position used in the last tests with these ball bearings, the results for Drag Turbine 4 and Drag Turbine 6B were almost identical to those shown in the table of the 10/05/2022 post.

I made a new sleeve for the drag turbine housing and started testing of Drag Turbine 4. I checked the bore of the new sleeve with a dowel pin that has a tolerance of +0.0000”/-0.0002”. The dowel pin required a light force to move into the sleeve without any apparent play. This is probably as close a tolerance as I can get on the sleeve bore and unfortunately the rotor still could tilt slightly in the bearings. I don’t know what the tolerances are on the dental ball bearings, but they appear to have a larger tolerance than the ball bearings I was using for the previous tests of the drag turbines. When I was adding and removing shims to find the optimum shim thickness, something occurred that I can’t understand. Because the maximum rotor speed was very low for each of the rotor positions I tried, I blocked the flow out of the exhaust port while the turbine was running to check for leakage. I expected the turbine to start slowing down and the pressure to go up if there was no leakage. Instead, the turbine speed went up. The maximum speed with the exhaust port open was 16,000 rpm and with the exhaust port blocked the maximum speed was 20,000 rpm. I haven’t been able to find a logical reason for the speed going up when the exhaust port is blocked and would appreciate any ideas of how this can happen. The following drawing shows the construction of Drag Turbine 4 and the only way for the air to escape with the exhaust port blocked is through a 0.032” diameter vent hole not shown in the drawing or through the ball bearings. The vent hole is in front of the ball bearing closest to the rotor and is intended to vent any gas that gets behind the rotor to keep it from entering the ball bearings.

I received the new dental ball bearings and tried them in each of my turbines that could use this size. All of the impulse turbines repeated the performance last shown in the 10/05/2022 post within 0.5 psi. My drag turbines performed worse than expected. This was the first time I tried the dental ball bearings in the drag turbines, and I found that the sleeve I added had an ID larger than the sleeves in the impulse turbines. I did all the sleeves the same way, so something must have slipped when I did the sleeve for the drag turbine housing. The oversize ID lets the ball bearings tilt enough to open the gap between the rotor face and the cover and allow too much leakage. I will make a new sleeve for the drag turbine housing and then try testing again.

Thanks for the feedback. I damaged the dental ball bearing by trying to pull it off the rotor shaft by gripping it around its OD. I was in a hurry to try another test and when I couldn’t remove the bearing gripping it with my hands, I tried gripping it with pliers. This was what damaged the dental ball bearing. The dental ball bearing shown below has worked extremely well for my tiny turbines running on air.

Jon Lawes
This dental ball bearing is used in the Star 430 dental handpieces that has a no-load speed of 400,000 rpm and operates at over a 100,000 rpm.

John P
The paper you gave the link for is very interesting and I plan on carefully reading all of it.

Thanks for the feedback. I damaged the dental ball bearing by trying to pull it off the rotor shaft by gripping it around its OD. I was in a hurry to try another test and when I couldn’t remove the bearing gripping it with my hands, I tried gripping it with pliers. This was what damaged the dental ball bearing. The dental ball bearing shown below has worked extremely well for my tiny turbines running on air.

Jon Lawes
This dental ball bearing is used in the Star 430 dental handpieces that has a no-load speed of 400,000 rpm and operates at over a 100,000 rpm.

John P
The paper you gave the link for is very interesting and I plan on carefully reading all of it.

To solve the problem of gripping the dental ball bearing mentioned in the last post, I turned down the diameter of the boss on Drag Rotor 2 to 0.180”. This is the approximate OD of the shims and is below the ID of the outer race of the dental ball bearings. The assembly drawing for Drag Turbine 4 R1 shown below shows this change and the added sleeve and dental ball bearings.

I added a sleeve in Housing 3 so that I could use the dental ball bearings in Drag Turbine 6B R1 as shown in the following drawing. After adding the sleeve, I added one dental ball bearings to Drag Rotor 6 without any shims to see if it could slide all the way down the shaft until it contacted the rotor. The dental ball bearing had a slight resistance the last bit of travel like expected, but when I tried to pull it back off the shaft, I could not grip the ball bearing tight enough to move it. Without any shims, the face of the ball bearing was tight against the face of the rotor boss so not even a sharp blade could get in between to pry them apart. I finally had to press the rotor off the shaft and then press the ball bearing off the shaft. After adding Loctite and pressing the rotor back on the shaft, and shimming the rotor to the correct position, I tried my first test. The maximum speed was way below what I obtained with the original all metal ball bearings so I must have damaged the dental ball bearing when I tried pulling it off. I ordered some new dental ball bearings and will start testing the Drag Turbines when I receive them.

Edited By Turbine Guy on 21/05/2022 20:12:51

I figured out a method to account for the compression of the nylon housing when tightening the nuts that hold the cover on and was able to get the rotor in the position shown in the drawing of the 03/05/2022 Post. This position gave me the best performance again. I moved the position of the rotor in tiny increments until I reached the minimum pressure of 18 psig required to turn the GWS EP 2508 propeller to a speed of 28,000 rpm. The required pressure changed from 20.5 psig to 18.0 psig in a total movement of 0.012” and from 19.0 psig to 18.0 psig in the final 0.002” move. This is much too sensitive for being due to space required for the gas to go supersonic since it is almost sonic with the 18.0 psig inlet pressure. This small distance from the face of rotor to the face of the cover probably affects the windage loss the most. Most of my sources say to close the gap between the faces of the rotor to the faces of the cover and housing to the minimum possible. The following test sheet is updated to show the latest test results of Axial Turbine 3N R2. Several runs were made with the 18 psig inlet pressure and the maximum speed was always approximately 28,000 rpm.

The following test results added the test of Axial Turbine 3N R2. I was only able to get the maximum performance in one run. I was in the process of trying different positions of the rotor when I got the lowest pressure required to turn the propeller at a speed of 28,000 rpm. I never was able to get the rotor back in the position that got the best performance. The nylon housing can distort enough when tightening the nuts that press the cover against it to close small gaps. To check if it was a bearing issue, I added the ball bearings back in Axial Turbine 3A. I was able to position the rotor as shown in the 30/04/2022 post very easily and the pressure required to turn the propeller at a speed of 28,000 rpm was a little lower than was shown in the test results of the 30/04/2022 post. I never had an issue with the nylon housing when running the larger gaps needed for the higher pressures.

Since Axial Turbine 3N already had a close clearance between the OD of the rotor and the housing I decided to add a sleeve and the dental bearings to it. The following drawing shows the position of the rotor that got the best performance. I updated the test sheet shown in the 28/04/2022 post adding the test results for Axial Turbine 3N R2 and it will be shown in the next post.