fluid and mechanisms - gears, joints, driveshafts, springs, dampers

MikeHersee

Hi, I'm looking for a solution to enable me to simulate the performance of a wave energy device. That is, a floating device that has a mechanism affected by the water. This means moving parts that are affected by water flows (for example, a turbine) need to be constrained to rotate about an axis that is itself determined by its position relative to other moving parts (eg, the vessel in which they are contained). I have seen DualSPHysics simulations that incorporate moorings so there is at least some degree of solid-solid interaction as well as bidirectional fluid-solid interaction, but does that interaction extend to, for example, univeral joints on a driveshaft?

Is it possible to have a floating water wheel that drives a pump attached to the same axle, for instance? Or an articulated floating vessel whose movement is constrained by springs and dampers. If both of these are possible, this would pretty much encompass all the technical needs. I have been looking through the documentation and at the moment I'm not sure if it could. I'd be grateful if someone with greater knowledge would let me know before I spend ages learning another package to find out it can't quite do what I need yet.

I know there are existing packages that can do this, but at a price substantially exceeding my current resources!

In appreciation,

Mike

Alex

Dear Mike,

I see your point.
I think it can be possible but it will require some numerical development....

By the way, what commercial packages you know that can simulate those devices?

Regards

Alex

MikeHersee

Oops, didn't see this reply - didn't seem to get notification of your reply. RealFlow is one such package, one of the very few actually. Even some quite high-end packages can't handle proper influence of mechanical objects by fluids, the solid-fluid interaction is one-way only.

However, I've now been exerimenting with RealFlow. It has five different methods for simulating water, but only one can properly handle floating objects - ordinary SPH particles - and that is the slowest method of all that uses only the CPU. It is extraordinarily slow, and the simulation I'm running right now only has 1 million particles to test some principles but is taking quite a few minutes per frame. In fact I was going to time it to say exactly but couldn't be bothered to wait for the current frame to finish. It's going to be unfeasible for the project I'm working on as I will need many more particles, probably closer to 10 million. For that reason I've returned my attention to DualSPHysics! Am going to post a more up-to-date comment though about springs and such like.

Mike

Alex

I have already replied to that post where I explained to you that incoming v4.0 will include the DEM source code and examples you can follow.

However, note that Realflow is not a CFD tool... it is only developed for nice simulations. On the other hand, I do not know now the current degree of maturity of Xflow however validations have not been published to prove the reliability of that code....

Regards

MikeHersee

I understand already that RealFlow is more aimed at visual effects than engineering-level precision. However, the learning edition, which essentially has the same functionality as the full edition, is less than £100, whereas XFlow is a 'contact us' price, which means a large number of figures before the decimal point! The idea is to get to the point where we can afford an engineering strength solution and the hardware on which to run it, so I'm forced to use cut price solutions in order to get funding. I'm sure many of us are in a similar position.

Regards,

Mike

Alex

Please, send an email to dualsphysics@gmail.com with the functionalities you need from DualSPHysics since we are going to release some of them and we are internally working on others that you may be interested. In that way we can find some kind of collaboration if you want.

Regards