Physics Help Forum Thrust of a biomimetic propulsion system

 Nov 2nd 2017, 05:57 AM #1 Junior Member   Join Date: Nov 2017 Posts: 4 Thrust of a biomimetic propulsion system Greetings to all brilliant and curious minds, Background: My name is Gio and I am working on a biomimetic underwater robot prototype. The biomimetic propulsion system consists in two lateral MPF-type fins (Median and/or Paired Fin) for loco-motionas shown in the photo. The propulsion system is controlled by a CPG (Central Pattern Generator) and the robot has also the ability to learn its optimum kinematic parameters for maximum swimming performance. It does that through a closed loop control system with a thrust sensor as the feedback mechanism. https://ibb.co/mJ7dDG The two fins undulate in a sinusoidal manner to create thrust as shown below. 9 servo motors, each with a max torque of 1.2N-m are used to make each lateral fin. Parameters of the undulating sin wave propagating along the fins can be controlled by the user. https://ibb.co/e63X7b Problem: The difficulty I am having is to create a model to predict thrust generation. I have conducted CFD (Computational Fluid Dynamics) simulations in Autodesk CFD 2018 and have data for the drag coefficient and force. However, as I am not a mechanical engineer, I find it very difficult to come up with a solution on how would one predict the generated thrust of such a propulsion system. Any ideas, thoughts or/and pointers would be highly appreciated. P.S. I am new to the forum so apologies in advance for any mistakes with the post. Last edited by Gio; Nov 2nd 2017 at 06:27 AM.
 Nov 3rd 2017, 06:33 AM #2 Senior Member     Join Date: Jun 2016 Location: England Posts: 883 The thrust will be dependent on the mass of water moved by the fins (per second). Is this information not provided by your CFD calculations? __________________ ~\o/~
 Nov 3rd 2017, 05:13 PM #3 Senior Member   Join Date: Apr 2017 Posts: 484 Interesting project ... The question is , why would you need a model to predict the thrust ?? there are only a few ways to drive this device , try them all and see which works best .... If you did come up with a model (not easy) you would have to test it against your practical results anyway ..., and probably find some factor was missed out . I can't thing of any large creatures that use this method for locomotion , nearly all move like a shark or dolphin, flapping their tail , this must indicate this is the most efficient system.... Although your method is probably smoother .
Nov 4th 2017, 04:01 AM   #4
Junior Member

Join Date: Nov 2017
Posts: 4
 Originally Posted by Woody The thrust will be dependent on the mass of water moved by the fins (per second). Is this information not provided by your CFD calculations?
- I guess it would be provided, but my difficulty is in modelling the deformation of the membrane of the fins as they move through water. It turns out it's not that easy to do.

Nov 4th 2017, 04:09 AM   #5
Junior Member

Join Date: Nov 2017
Posts: 4
 Originally Posted by oz93666 Interesting project ... The question is , why would you need a model to predict the thrust ?? there are only a few ways to drive this device , try them all and see which works best .... If you did come up with a model (not easy) you would have to test it against your practical results anyway ..., and probably find some factor was missed out . I can't thing of any large creatures that use this method for locomotion , nearly all move like a shark or dolphin, flapping their tail , this must indicate this is the most efficient system.... Although your method is probably smoother .
There's a couple reasons behind thrust modelling:
- understanding the hydrodynamic characteristics of the prototype
- I may/may not want to publish my work which requires creation of a CFD model as well as experimental data

 Nov 5th 2017, 04:19 AM #6 Senior Member     Join Date: Jun 2016 Location: England Posts: 883 Ah, yes I see your problem. Most CFD solvers provide a single solution for a static configuration (unless you start spending serious money and computing power). However you can probably define some limiting conditions like when does the flow become (unacceptably) turbulent. I would suggest that the force on the fins at this speed will give you an indication of the maximum thrust you can (efficiently) apply. I would expect that you would find the force will rise quite steadily with speed until a quite clearly defined point when the rate of increase of thrust with speed will drop off sharply (equivalent to a wing stalling). Find this point for a few static configurations, modelling different points in the sinusoidal cycle, and that should give you an indication of what your maximum thrust might be. PS I am fairly sure that I have seen a nature documentary which showed a (small) fish that used this form of locomotion. But I cant dredge up more than a vague recollection. __________________ ~\o/~
Nov 5th 2017, 04:59 AM   #7
Junior Member

Join Date: Nov 2017
Posts: 4
 Originally Posted by Woody Ah, yes I see your problem. Most CFD solvers provide a single solution for a static configuration (unless you start spending serious money and computing power). However you can probably define some limiting conditions like when does the flow become (unacceptably) turbulent. I would suggest that the force on the fins at this speed will give you an indication of the maximum thrust you can (efficiently) apply. I would expect that you would find the force will rise quite steadily with speed until a quite clearly defined point when the rate of increase of thrust with speed will drop off sharply (equivalent to a wing stalling). Find this point for a few static configurations, modelling different points in the sinusoidal cycle, and that should give you an indication of what your maximum thrust might be. PS I am fairly sure that I have seen a nature documentary which showed a (small) fish that used this form of locomotion. But I cant dredge up more than a vague recollection.

Thank you Woody, in fact this is what I am doing right now. However, I would prefer to be able to model the dynamics of such a propulsion system more accurately. Meanwhile, I found some other software (FLOW 3D) that could allow me to do so, will try it out and see if I get better luck. Also, simstudio + openfoam seems to have better capabilities than Autodesk CFD, but I have never used it. In any case, I will consider this case solved as I know where I am heading from here.

Best of luck everyone.

 Thread Tools Display Modes Linear Mode

 Similar Physics Forum Discussions Thread Thread Starter Forum Replies Last Post Simonamf Thermodynamics and Fluid Mechanics 16 Oct 27th 2017 12:08 PM avito009 Physics 2 Feb 13th 2017 06:27 AM crowxe Thermodynamics and Fluid Mechanics 0 Apr 23rd 2016 06:30 AM arcana Equilibrium and Elasticity 1 Mar 25th 2016 04:00 AM chocolatelover Kinematics and Dynamics 1 Nov 16th 2008 01:20 PM