Just google "DIY flow bench", masses of info on the net about it. Measurement is as simple as two fluid manometers i.e. water in a clear plastic tube!

Check out dtec.net.au for a good tech article on a DIY flowbench.

I'll try amd dig up a link - something to do with tractor pulling - they had a whole forum dedicated to diy flowbench builds, initially inspired by an old article in hot rod or car craft (something like that - a US published big v8 musclecar type mag).

Anyhoo, what they essentially do is have enough vacuum cleaner motors to generate airflow (and the newer twin fan ones are actually better, as apart from flow they can actually develop more vacuum/pressure drop ) in the range required, some plywood cabinets, more or less with 2 major chambers - one in one out (or perhaps it would be better to say one pushes for the exhaust, one pulls for the intake, sortof), then hooked up to a third chamber which is routed through a large enough area to allow flow to straighten out etc (so it doesn't distort the true pressure drop and hence the flow measurement.

that chamber is then routed to the table top area, which will haqve a cylinder on it (for best results you need a cylinder with a bore that matches the engine bore of the engine you are working on, otherwise you can't factor in valve/bore wall shrouding which dramatically affects flow. It's one of the dodgy porters 'tricks' to make it sound like they got more flow out of a head. Ironically this came up recently when a bloke on another forum enquired about a head advertised for sale elsewhere,with some vauge info about what it flowed. Some of the discussion was via email, so I can't divulge all, but suffice to say the head was flowed with a 4 inch bore on there (holden was 3.625, and worse still, the holden heads don't sit over hte middle of the bore, instead the valves are off to one side, so they have significant shrouding issues. Point being the advertised head wouldn't flow anywhere near as well if it was tested on a proper bore sleeve adapter that positioned the head as it would be on the engine.

but getting back on track

then you have a 'selector' which is nothing more than a steel plate with different sized holes cut in it (very specific and accurately cut holes). Basically through a fixed orifice with flat/sharp edges, at a particular pressure drop there will be a particular flow, so you select the hole (to let the air flow through) to suit the general flow range you want (which is simple enough to work out, you'll be dealing with less than 300cfm on intakes 4 cyl heads almost certainly SO they draw the air through the fixed sized hole and through the port n top of the cylinder adapter (I mentioned bore size was important, I'd also add that for best results to relate to real world operation, pretty much people have found that if you make the cylinder at least 1.5 times the engines stroke in height it seems to be best. Smokey Tunick actually pioneered this before we were born, by flow testing with the head on an engine block and pulling the air out through a custom sump with the other valves/bores blocked off, as well as pioneering about a million other things we all take for granted these days.

OK they pull the air through hte orifice and the porr and use a 'tap' (which varies how much air can be bypassed through the cabinet and how much has to come through the port, - a bit like a wastegate flap or a throttle, but bigger) and that is adjuted till you get up to the test pressure drop (25 and 28 inches of water pretty common ones. That reading will come off the 'main' manometer. then there is a differential manometer, on a diagonal (makes it easier to read/more accurate, sort of ) and that compares the pressure drop with the one the orifice creates and the valve/port creates. If the head flowed as much as the orifice then the manometer would be on the 100% mark, and if less, you can baically calculate the actual flow - it's a comparative flow device if that makes sense.

Believe it or not these home built jobs can be remarkably accurate. I've mentioned before that even commercial flowbenches will vary from unit to unit a little and sometimes on different days if barometric pressure is massively different but tested on the same general type of weather/temp (might be a good argument to have one inside an insulated section of the garage that is more even in temperature. Anyway, whilst there might be a few cfm different from machine to machine, they are accurate to themselves in similar weather, so you can easily track the changes.

Another option for moving the air is believe it or not an old 'big' supercharger (something like an 8 71 can flow well in excess of 600cfm at attainable rpms (a few thousand rpm, with sufficient gearing and an electric motor - probably 3 phase by that stage, but of course one can go smaller. FWIW the toyota superhcargers are a little small for this) There are also industrial fan motors which can move decent cfm and pressure as well (be very careful with some fan setups as they flow well freely, but can't flow much at all against a vacuum or pressure.

Aido - if you have access to the valve grinding gear - presumably throat cutters too (not that that is necessarily needed for a mild port job) - maybe someone can get back with the port cros section area (from which you can get a bit of a read on what valve diameter and throat size will work with it. I can probably give some rules of thumb (which i definitely didn't come up with myself) but it'd be even more helpful if BigE or someone who eats and sleeps datto heads can chime in (because theory aside, nothing beats what is called engineering science - which is essentially understanding the basic needs but then testing relentlessly to get the best flow and then using that to work out why it works the best.) i.e. good old trial and error (well a guided sensible approach to trial and error).

On the seats, there would be more flow in a 5 angle job, or if you were massively dedicated, the only 'discrete' angle would be the seat contact faces themselves, and the optimal flow is to hand contour the rest of the angles for a better transition (the gain over a 5 angle is incredibly small, so I'm only mentioning it for trivia sake), don't go too far on the short turn radius (from what I've seen of them, they are far from the worst factory example out there). Probably the 'big' gain to be had is the valve guide/boss area. you can trim around there pretty well.

A while back I did a quick MS paint diagram, using a pic someone uploaded (I forget who sorry) - I altered it to show where material could be taken away without any fear of losing flow. Since then I don't have any webspace, so I've re-done them a little bit and put in a photo album:








Essentially pop the valve guides out the way, and open up around them. On the intake, bullet nose/taper the tip of the valve guide. it still retains the same length just thinner toward the edge. On the exhaust side, don't bullet nose the guide (and also even though flow is improved leave the exhaust valve/seat contact width thicker for better heat conductivity so it runs cooler, it's worth it to trade a little flow for a safety)

Cut away the red areas. pretty self explanatory. I've tried to show the sort of short turn to go for, basically a little more rounded, but you don't have to go nuts there. on the intake port, if you wanted to increase the size of the port in general, then you take the material (whenever possible, and in this case it is) from teh roof, not the floor, makes it ever so slightly more downdraught less of a sharp turn into the cylinder. I've shaded that in blue - but it's totally not essential, the main gains will be removal of the red shaded areas.

I'm not sure if it shows up on the pic, but very generally on the intake you try and make it slightly bigger at the entrance tapering down narrower toward the valve. Not massively - very slight - like around 5-7 degrees which probably works out to something tiny like a few mm (maybe 5 bigger diameter at the manifold face and another 5mm bigger again by the time it gets to the base of teh carbs (for twin carb setups). Again, not essential, but worth a look.

If the port is smaller than the intake manifold then open up the port till it matches the runner, or there is a bottle neck that compromises flow and can see fuel dropping out (or alternately create turbulence which keeps it suspended but costs way too much flow to do it) If the port is already biggerr than the intake manifold opening, then you can open up the intake to match it, or alternately open it up till it is just 1mm less diameter. the mismatch in that direction doesn't hurt flow, but it can handily help resist flow reversion which can happen with a big cam at low rpm, because of how late the intake valve closes. On a similar note, if possible, seat the exhaust valve just a touch higher in the head when cutting the seats (I don't mean by a mile). It will put it closer to needing new seats down the track, but having the intake seat a little lower (closer to the piston, and the ex a little further away, on the tail end of the exhaust cycle and the beginning of the intake, i.e. valve overlap that slight mismatch can help just a touch with exhaust going out, and pulling some intake in, without exhaust reversion. If you hypothetically dragged a trace amt of air/fuel out the exhaust, it'd actually be a good thing for power, as it means a cleaner intake charge, no exhaust contamination, which will mean more power than the trace amount of fuel being lost but a compromised burn quality.

Last thoughts for this post (sooner or later I'll find a cure for insomnia, or maybe just move somewhere so i can work on the datto late at night) - it won't improve flow (well not airflow anyway) but one of the things done on a few dinosaur engines, that could be applied here - with all the valves/springs etc stripped off teh head, polish the top of the head everywhere inside the rocker cover (except where rockers bolt on and valve spring seats etc) - remove the casting finish and get it as shiny as possible, and then polish the oil drain back areas, radiusing and smoothing out wherever applicable. This will help the oil return to the sump easier, which means you'll not (potentially) get to a point it isn't returning quick enough and the sump is startng to empty. On some other engines you can lose oil pressure after sustained racing speeds, and if you are quick enough, can pull over, pop the bonnet and pull the oil cap and see the oil all trapped in the rocker cover (of course good breathers also help here as does a good ring seal so that blowby isn't trying to all escape through the rocker cover breather) I certainly don't consider the datto a series to be in this category, but by the same token, there's certainly no drawback to such polishing work.

Bottom line, it'd be pretty hard to get into trouble going along that basic path. It might not get you on the front row of ipra or 2 litre sports sedans (hell that's a given) but it's not a bad start.