Roman

Going from 2nd to 3rd throws away some of my powerband, why would I want to do that.

My idea: Leave it as is Your idea: Remake everything. Yep sounds difficult

Yeah I've seen that vid, its easy enough to vary the flow between runners like that with a time dependant analasys. But yeah there are too many variables and assumptions to use it for anything apart from cool animations. But as per my rant above, knowing the MAF sensor value in grams per second gives me some useful info regarding what's good for anything upstream of the throttle body. It's a real life value which can be applied to known pipe sizes etc. Rather than making assumptions about how much flow/volume/blah blah comes from the valves/cams/etc. which is what's required to reinvent the intake manifold which is already good. Everything I've done so far (not much) seems to point to that the bigger the pipe the better, for minimising pressure drop. Which is useful info. Rookie - As I've mentioned briefly before, my car used to have quads on it. But to get a runner length that's good with a straight pipe, I had to cut some of the bonnet support things out. This caused an irrepairable bonnet warp and I had to source another one before painting the car. Having a bonnet cutout or whatever isnt an option for what I want with the car. If the car needs bent intake runners then the only real advantage of quads over single TB is that you can have a much larger plenum volume and no step down to the diameter of the single TB. Which is beneficial, but if I had $2-3k to throw around then I'd be buying tires and other things first.

Hey cool article! Thanks

Oh so as per above. It's obviously a good plan to put an air intake into a high pressure zone. I've Been wondering if there's any benefit to making a 'trough' type thing, that collects and holds air slightly pressurised for the intake pipe to pull from. If the amount of air coming is greater than the amount of air going out then it should build some pressure. So... If you've got a certain mass of air flowing out of a hole in the back of a box/trough, (known value as per airflow meter readings) then if the trough is still lower than atmo pressure when the car is going 100kph then you just need to keep making the box bigger. So I started out with a box 100x100 with an outlet that's sucking 170 grams per second of air. Tested at 100kph... Lower than atmo pressure in the box. Made it 100x500. Still lower Made it 100x1000. Now gets to about atmospheric. (This could fit in the car, could take up the whole lower inlet panel thing) I then drew some piping and an airbox type thing off the back of it, to represent air filter box. To see if that could get slightly above atmo. Yes and no, the trough thingy gets above atmo but having a 72mm dia hole outlet is where the pressure drop happens. Increased the size of this, and increased the size of the box to 200x1000 Now it's getting above atmo pressure inside the airbox at 100kph... However it was above atmo pressure by something like 0.001 PSI. Haha. Not bad I suppose if it means you could get the entire inlet tract including air filter for 'free' in terms of restriction. It seems the name of the game is to make any pipe pre-throttle body as bigger diameter as you can get away with. In my case I'm stuck with a MAF sensor that needs to be inside a certain diameter pipe and it's got a 72mm throttle body so those seem to be the bottlenecks. I wonder what the intake pressure drop is at the moment IRL... I've really gotta get a pressure differential gauge at some point for nerdy experiments like this. I guess another thing is that 170 grams per second is peak airflow meter reading, but because the engine operates through a powerband it's only reaching this at certain speed/rpm combos. It would only get to this reading currently at 50kph, 80kph, 130kph, ~185kph. So at the real life 100kph MAF reading it may very well make positive pressure there. Even if it only did so 'off peak' it just means it would broaden the powerband instead of increasing the peak. I might run it a a few more times at the grams/second rating for the rpm it drops to on gear change. And yes I realise this all seems stupid compared to just getting a turbo or supercharger. However engine bay simplicity is a set in stone goal for my car. If I made a contraption as per above it fits into a part of the car that's currently unused and completely out of the way of everything else. (and doesnt cost me any money hah)

Yeah it's got the modelling stuff. Jeepers, Ansys looks pretty hardcore. Might check that out another time but just learning solidworks is enough to fill my brain to the brim currently! And my knowledge of engineering type stuff is the bottleneck anyway, not the software. I'll put that on the to-do list!

Hey so one thing I've been wanting to play with in solidworks is some of the flow modelling stuff. For sake of learning the program I thought I'd draw up an intake manifold / throttle body / piping / air filter / air scoop / etc. Then play around with some stuff and see what changes. For my motor I know: Intake runner lengths, diameters Approximate plenum size/volume Throttle body diameter piping diameter How many grams per second of air the motor currently consumes at full rpm/throttle with the current length/size/etc of inlet tract (as per airflow meter readings) So what I've been thinking is that it's probably outside the scope of what Solidworks can accurately acheive to try and model the pulses as each cylinder sucks etc. As although you can make time dependant flow it's not going to accurately represent an engine well enough. So what I was thinking is that I've got a known Grams/second at a known pipe size (start of airbox) so perhaps I could model some different intake configs upstream of this (as realistically I'm not gonna build a new plenum or anything anyway) and see if I can minimise the pressure drop. Or, since ideally I'd want to acheive a higher grams/second perhaps I could set an 'outlet' as vaccum or whatever at different strengths until it shows 170 grams/second as the mass flow rate with an approximation of my current intake there. It will be easier that way to see changes as mass flow rate seems a lot less like reading tea leaves than pressure drop which seems less sensitive(which I guess is why they use MAF in modern cars so much) and also obviously increasing the mass flow rate is the end goal which gives an engine more power. I've been thinking about whether a big bathtub looking scoop thing in front of the front radiator panel could build positive pressure past the filter at 100+ kph.... Making the intake tract up to that point 'free'. I would probably need to model a reasonable amount of the front of the car but thats fine because the point of this exercise is to learn how to use solidworks more than anything else / cbf actually working on my car haha.

The only problem I can see with this is that if all four wheels are spinning, it's going to interpret this as though you're travelling at a higher speed, and so increase the boost and then your wheel spin increases more. Or if just front or rear wheels are spinning... How does it know which ones are or arent? (Is this for Sentras car you're talking about?) Limiting per gear is probably a better option? If you can look at the datalog and figure out the first rpm/load combination that the wheel spins at. Figure out how much HP this is. Then apply the horsepower formula to the rpm etc of the wheel to figure out how much horsepower the wheel can put to the ground at various speeds then limit your boost to acheive approximately that amount of horsepower at each speed.

It's a 20hz chipset bodged together with an arduino or something something that my Dad pieced together. It's a Skytraq Venus something something, I think it's an older model now as the ones you get from them are now 50hz. All I know is that it plugs into USB and is damn accurate! With the program Racechrono on a netbook.

I think I'm gonna have to bring some warm beer and see how your wizard machine works some time Dave.

Steam engines run on heat, and in the first 'pump' they havent used it all up. So some have 2 or 3 (or more?) chambers where it goes from first, to second, to third. So the air that comes out the end of all that doesnt have much energy left. By the 'end' of steam trains they had them figured out pretty well. This is relevant to my interests, how are you making the point cloud in the first place? Is that with your fancy laser scanner thingy?

Before I had GPS, I used to go to a trackday. Drive some random lines. Someone would be pointing a camera at the start/finish line and rolling all day. A few weeks later, someone would tell me "Oh hey, on the 3rd lap of 2nd stint you did XYZ time" Which is cool, but effed if I know what I was actually doing then? With datalogging GPS. I can go do 5 different laps taking 5 different lines. Come back into the pits, and it will show me the best way out of those to take each corner/set of corners. Then 5 more laps running slight variations of that best combo. And so on. The improvements you can make in just a day are amazing, if you're learning a new track or whatever. Especially when you're trying to find that trade off between the 'racing line' and minimising the distance travelled to complete a lap. Some of the biggest gains I've made with a GPS at Taupo have come about on the realisation that I was making the track about 20 meters longer than it needed to be... By taking what I would have thought were good lines that give good exit speed. Highly useful!

No way man! I've driven on semis before, the difference is astounding. I've already got a 20hz GPS logger + racechrono which is pretty good for pointing out my shoddy driving

Haha, stuff like that is firmly in the day dream category - as you say there are many other things which would help more. Semi or full slicks would be best $ per time improvement at this stage. And wont likely have any of those any time soon, need to ease up on car spending for a while. And the car's at the point where it handles okay and getting more practice and having some fun is best $ spent. The GPS datalogging and engine datalogging I've got at the moment is enough to keep me happy. And these datalog using my laptop and tablet, which are both something I can justify owning for the 99.999% of the time that I'm not in a car on a race track. Dedicated datalogging stuff is a bit too single use for my dutch blood.

Yeah I am keen to view my sloppy gear changes and ham fisted steering in excruciating detail.

That was a different kind of video I think. Slightly related but I've been thinking it would be interesting to setup some potentiometers to datalog suspension travel etc. Live axle makes things easy at the back, haha.

NicT can answer that one. I remember seeing him digitize some sneakers or something haha. Or was that from photos. Cant remember.

Okay I'l see how I go.

Yes but that plane could still rotate 360 degrees while being perpendicular to the end of the pipe, if that's the only thing defining it.

Yes! But then what stops it from being perpendicular but rotated 90 degrees or whatever? Does it depend on the plane you start in.

I used to draw 3d models of buildings for like 7 years, but in Autocad. Any time I've just needed basic flanges etc drawn I find its way quicker in Autocad. But want to learn solidworks to play around with the flow/temperature/whatever kind of stuff.

If I select the round face of a pipe end then there isnt enough info to show what the plane needs to be perpendicular to? Do you use solidworks much?

Puhuhuhu

In other news, I've been learning how to use solidworks again. But I'm getting a bit stuck with the UCS I can figure out how to make a new plane on the face of something, but how do I make a new plane that's say 90 degrees rotated from that? I'm sure there's a simple answer but it's just a kick in the balls coming from Autocad to Solidworks.

Suggesting a K swap for everything is an idiotic one dimensional opinion that doesnt suit anyone without excellent fabrication skills, lots of time and/or $20k to spend. I am not in any of those 3 categories.