9 Replies Latest reply on Aug 6, 2014 12:10 AM by Jared Conway

    Diverter Valve Simulation- Expertise Needed

    Brandan Philbrook

      This project of mine has many improvements to be made, but the main one I wan't to focus on this post is the Simulation/Motion study. I wan't to simulate the drive process to predict the life cycle of the diverter with the indexing cyllinders being made of Alloy Steel, which impacts the ASTM A36 Steel Sprocket. I then want to choose a more suitable material for the sprocket to increase the life of the diverter.


      If this is not possible, I am currently trying to animate the process, and am finding it difficult to set the indexing cyllinders so they will not penetrate the sprocket and instead rotate it. In the attatched file, I included a working analysis of my diverter. ANY help very much appreciated


      Below is more information about the 4-Way Swivel Diverter that I modeled and machined from scratch. This diverter is used in various pneumatic conveying systems. Product flow goes into the single pipe, and out through one of the four pipes at the opposite end. The plate can be rotated (clockwise direction only) by a patent pending drive process 




      Here is how it is controlled:




      Below is the drive process:



      A little more detail of operation:


      1. The cylinder action is reversed by an electronic air solenoid (not shown), and the air cylinder pushes the rotary plate out (separating the sealing surfaces on the top and bottom of the valve).

      2. When the rotary plate is pushed out, the two air cylinders on either side of the center shaft cylinder push the clutch disc into the rear of the sprocket to hold the entire rotating assembly from moving as a result of any imbalance in the rotating assembly.


      3. The first bank of the miniature indexing cylinders (x4 of the x8 total) actuate and impact the sprocket and cause the sprocket to move. However, the moment the miniature indexing cylinders actuate, the clutch disengages the rear of the sprocket leaving it free to rotate by impact from the miniature indexing cylinders.


      4. After the first bank of miniature indexing cylinders actuate. The clutch re-engages the back of the sprocket to hold the rotating assembly in position while the bank one indexing cylinders retract. Once the bank one index cylinders retract, bank two indexing cylinders actuate, pushing the sprocket further around. Again, at the same time the bank two cylinders actuate, the clutch disengages to allow rotation.


      5. After the bank two cylinders actuate, the clutch re-engages to hold the rotating assembly in position. Bank two cylinders then retract.

      6. Steps 2-5 repeat until the valve has moved to the desired port. The geometry and design of the drive sprocket allow the valve to rotate 30 degrees for every one cycle (actuation of bank one and bank two) of actuations.



        • Re: Diverter Valve Simulation
          Simon Yang

          Very interesting!

          Sorry. The attached files can not be opened as no any references here. Please use pack&go to upload the files.

          • Re: Diverter Valve Simulation
            Jared Conway

            very ambitious


            my suggestion is to break down your needs and simplify what you need to get done

            if you want to show how it works, use animation

            if you want to understand how fast it rotates, what kind of torque is required..etc, work on a motion analysis, but spend time simplifying the assembly first

            if you want to choose a differnt material for a specific part, don't try to analyze the whole system, figure out the loads on those parts and run the analysis on those parts on their own

              • Re: Diverter Valve Simulation
                Brandan Philbrook

                I want to run a fatigue-type test that determines the wear on the "softer" sprocket over time due to the impact of the cylinders. The cylinders actuate of a force of 68 lbs at 100 PSI input pressure.

                  • Re: Diverter Valve Simulation
                    Jared Conway

                    i saw your simplifed analysis

                    wwhy would you not do this analysis with just the gear and forces to simulate the actuation

                    you'd need to go to nonlinear to add the "sequencing" but if they aren't happening at the same time, wouldn't your analysis boil down to a simple application of one actuator?

                      • Re: Diverter Valve Simulation- Expertise Needed
                        Brandan Philbrook

                        The spacing of the actuation causes the sprocket to be rotated at different teeth every revolution. Is there a simulation study where I can, well, simulate the force on the sprocket at each instant, rotating the sprocket 15 degrees every actuation? My goal is to determine how long the sprocket will last in the diverter with different materials. Also, a stress concentration generated model would be ideal so I can adjust the spacing and contact locations of the stresses on the sprocket. Lastly, a motion analysis to properly show how the process works. 1 full revolution of the sprocket preferred (24 actuation's)

                          • Re: Diverter Valve Simulation- Expertise Needed
                            Chris Michalski

                            If you simulate the force at one instance, it will be repeated at the other faces.  I can understand that for a fatigue study you want to take into account cumulative damage at adjacent teeth instead of simply repeating the load on a single set of teeth.

                            But I would start with simulating a stationary force application.  The teeth are most likely to break locally due to point loading, or the entire tooth break off.  Neither of those is likely to be largely dependent on the cumulative stress in the neighboring tooth so applying indexed forces is probably of minimal benefit.  Splitting the hub (which could be more affected by adjacent cumulative damage) looks very unlikely due to the vast difference in cross-section vs the teeth.

                            My thought would be: start with a static simulation of actuator force against the tooth.  Then run a fatigue study of the same.

                            I would almost bet that the effects of sliding friction (erosion and wear) are going to affect operation more than pure strength of materials and fatigue cracking.  You'll end up wearing the teeth to the point that your motion isn't the same.

                            1 person found this helpful
                    • Re: Diverter Valve Simulation
                      Brandan Philbrook

                      Thank you for your responses.


                      I have simplified the system to:




                      • Cylinders 1,5 and 3,7 actuate at the same time
                      • Cylinders 2,6 and 4,8 actuate at the same time
                      • See the attached drawing

                      Where the materials of the indexing cylinders and the sprocket are set. Attached is a folder with the design study assembly, 3 components to the assembly, and a drawing with dimensions of the spacing for clarification and actuation detail. Let me know if this helps