Technical Progress Report After Tilt Test Dummy Trials

So, What’s It All About, This VR Driver Simulator Training Thing? Forums Prototype Technical Progress Report After Tilt Test Dummy Trials

This topic contains 0 replies, has 1 voice, and was last updated by  Vince Sunter 9 months, 1 week ago.

Viewing 1 post (of 1 total)
  • Author
  • #20764

    Vince Sunter

    Tilt trials now completed with 120kg “King Jack” test dummy and they have been a raging success. Well, except for the broken left side mirror I missed on earlier range testing but you can get with full tilt back plus turn right.

    The purpose of these trials was to determine optimum system pressure and set the accumulator to 90% of this pressure so it can function as a viable capacity increase for the pump. Plus see if the planned speeds could be achieved or if the accumulator might be needed to get us there. The required motion speeds were based on real world driving experiments as to how vehicles behave in the range of typical situations including emergencies.

    In every case, better speeds were achieved than what I had decided was necessary. For acceleration and braking max speed is almost double that needed. For turning left just a bit more and turning right capability is nearly double – this differential is due to how cylinders work not any other “special stuff”. In essence, going one way has to fill the entire cylinder whereas going the other way the support rod reduces the volume about 40% so it has less fluid volume so fills it faster. There are dynamic effects going on too, but you don’t need to know about all the subtlety of that!

    These tests also revealed what I suspected, namely that the bounce cylinders have a smaller job to do so less pressure needed – hence whenever they were being extended at max rates the other cylinders couldn’t also extend, although they could retract – at around half speed. If the bounce cyl speed was reduced then more control existed for the other cylinders. Adding the accumulator should largely eliminate this issue, but throttling bounce cyl operation is also a possibility.

    I also note my original logic on the bounce cylinders was wrong too. They have an open centre, so as soon as you stop lifting them they return via gravity. I had thought when they were near the bottom I could activate them to run the return through the throttling valve so the last bit would fall at a rate controlled by the valve. In effect, creating a cushion for the return stroke. When you look at exactly how hydraulic equipment is arranged, this seemingly simple thing is not going to happen! So I may go back to a closed centre valve spool and operate bounce around the mid-point. Which will make garage ceiling clearances all the more important!

    The Arduino programming set-up will have a table of “no-go’s” for all the possible cylinder location combinations, so that will be fine. For now, I have to miss, or I smack holes in concrete, or knock off mirrors it turns out. Thank goodness for Araldite!

    One thing was evident is how over-designed the structural elements are, from a strength perspective at least. 120kg of sand-cement test dummy was “nothing”. I am climbing all over it with an extra 90kg as dressed and no hint of more than a mm or two of deflection. Even when I go to the furthest corners and bounce on them not much happens. I smacked it into the concrete a few times and it is like swatting flies – the swatter is always going to be fine and the flies get smashed. But conversely, when making “extreme movements” some macro flexibility was noted. This was mostly from the bottom leg underneath the car – which may need some stiffening at a later date, if it is an issue during more normal movements.

    Next steps are to get all the gauges installed in the control panel, plus replace the potentiometers with the higher rated ones. And of course get the accumulator charged and repeat the system trials to see what changes have occurred. Then can start the serious interfacing work to make it so the hydraulics are directly driven from steering, brake and accelerator input via the Arduino controller. Then sort force-feedback for steering. Then do the PC interfacing work and get vision fully matched to visceral movement and finally real world vision overlapping. Simple eh, when you list it like that – let’s get cracking!

Viewing 1 post (of 1 total)

You must be logged in to reply to this topic.