Technical: Duratec ports


If you own an MX5 derived 2.0 engine and aren’t interested in all this technical gubbins then you just need to know you’ve got the one with good flowing ports; they never put a bad head on a 2ltr NC engine.

If you’re reading this as a non MX5 owner or are doing an engine swap and need to know more about heads then make yourself a brew and read on, I try to make it easy to understand for the average Petrolhead, but it’s difficult for me to judge the possible gulf between my knowledge and yours so bear with me. Similarly if you want to know whether to get your head ported or not then we’ll get to that later too.

Studying where the air goes inside a flowing port. The smoke leaves a residue behind so you can actually see afterwards where the flow is concentrated.
I usually do this last thing before going home, it fills half the workshop with smoke!

As far as the N/A (naturally aspirated) engines are concerned all the exhaust ports, exhaust and inlet valves are the same, however there are 3 different shapes of inlet port. It’s been debated for many years, but no-one has ever gone publicly into too much depth about it.

Whenever I encounter a new head that i’m going to be doing a few of I study it in quite some detail and spend many hours with it on the flowbench getting to know what works and what doesn’t. A flowbench is just one of many tools which help an engine builder create a picture of the engine because once you’ve understood it, then you can start to modify it.

The bench does more than just give you a CFM figure, with the right additional equipment you can also check velocity, flow bias, disruption etc. Velocity yeah, how many people consider how important that is whilst waving their cock around with their big CFM figures as if that’s all that matters?

Real porting is about changing the shape of the ports, it isn’t about smoothing and polishing the walls. For various reasons the engine does not want smooth polished ports, it needs a rough finish. Think about it, the engine manufacturer sets up a mould to cast thousands and thousands of heads from. If they wanted a smooth port finish they would have made it smooth, but they didn’t. Anyone who polishes an inlet port is clown.

I’ve been doing head flow development and porting heads for many years now, so much so I’ve got nerve damage in my fingers from holding a grinder for so long. The first time I saw a Duratec head I thought yeah, they REALLY knew what they were doing here, it’s an automotive work of art. All the features we’ve been creating as much as possible to old heads over the decades are there already.

Perfect short side radius? Check. Well formed splitter? Yep. Min CSA ratio matches valve throat? Of course. Strengtheners in the bowl to prevent cracking on an overheated head, D shape over the SSR, 3 angle seats, lightweight valves with thin stems, a hump in the middle of the SSR to divide the flow around the valve stem, well formed bosses cast around the valve guide; the list goes on. Someone with a CNC mill or a hand held grinder will go in and plough all these features out in the wink of an eye. The valves will flap around in the shortened worn guides that they couldn’t be bothered to replace, never seat properly and wear prematurely, but because they smoothed everything out, got 20 likes on FB and told you they’ve put a 3 angle seat on you’ll think they’re great.

Here’s a thing; if I took a grinder (only) to a port, re-shaped it here and there and bolted it onto an engine we’d see a power increase. If your average customer saw it they’d say ‘wtf is that i’m paying for? It looks like a dog has chewed it’. The customer thinks he knows what the engine wants; lovely smooth polished ports. The engine on the other hand doesn’t give a toss what it looks like, it just wants airflow and properly homogenised fuel/air. So we finish the ports off to satisfy the customer and charge them accordingly, they post pics on FB and get 30 likes, job done.

Port molds are one of the tools I use to get an idea of what i’m looking at, If you don’t spot relevant features simply by looking into a port, you’ll see them on a mold. Here are the various parts of a port labelled, but I generally refer to the ‘short turn radius’ as the short side radius (SSR) and the area opposite it below the valve seat (shown here as long or back wall) the bowl.

What isn’t marked on there is an important feature called the minimum cross sectional area or Min CSA. Also known as the choke point this is one of the main governing parts of the port, the other being (aside from valve lift) the valve throat, remember this, it’s very important.

The Non VVT 1S7G and 3S7G head found on low powered Duratecs such as the Focus, all 1.8 engines and some Mazda 6 are to be avoided if a good port shape with equally good flow is desired, it cannot be improved upon to the same shape as the 3S4G (2.3) 6M8G (2.0) & L5 (2.5) which are found on the VVT engines. The ports have been given the label ‘High port’ and ‘low port’ years ago and we’ll stick with that, no point in trying to change things now and it’s difficult to describe them in a handy to use term as you will see below. The reason it doesn’t flow as well is quite apparent when you look at these molds. Low port is on the left, high port on the right.

This picture is a touch deceiving as the port on the left looks shorter and of lower angle, but in reality it isn’t. If you look at the line where the valve seat ends you’ll see it’s canted over a bit.

As you can see the low port has no SSR or bowl. These two features are solely responsible for its lack of flow as they are very important parts of a port and help turn the air. I haven’t spent much time on it as it’s a non starter as far as power is concerned, but it looks like a tumble port which is good for economy and low rpm performance.

When designing or porting a head there are a lot of criteria to be met and the unwary can really mess things up with a grinder, but one thing stands out over all the rest and it is the fact we are trying to get the air to turn a corner losing as little mass flow and velocity as possible.

You can liken this to a car taking a corner at speed, if you try to perform a sharp 90 degree left at high speed you won’t make it and the car will simply skid straight on. If the angle of the corner is lessened, lets say to a large radius or two 45 degrees then you stand a better chance of making it round at the same speed. Air flow behaves in much the same way – the greater the radius, the more chance it has of hanging on and getting round without undue turbulence which slows it down. Now you know that apply it to the picture above.

Taking the above into account there is something else to be wary of. That is people who think they can flow test at 10″ of test pressure and scale it up accurately to 28″ or more. Air is heavier than you think, if it makes it round the SSR at 10″ it may not at higher test pressures when it’s flowing faster. This is quite easy for me to prove, I can test at 10″ and scale it up, or ‘correct it’ to 28″ (multiply by 1.670), record that figure. Then switch some more motors on (I have a 6 motor bench) and actually test at a real 28″. Do the new test figures equal the calculated figures? No sir they do not. Do major car manufacturers test at 10 or 28″? Nope. The higher the test pressure, the closer to real life you will be and the more accurate your figures will be too.

So what of the 2.5 head?

2.5 port on the left with wider entry (only).

Often touted around as the one to have and has massive ports, huge flow you’ll be surprised to learn it actually isn’t all that and many people were duped into buying one. Some people will even tell you it has bigger valves! Only the port entry is larger, but it soon reduces down to the same SSR and Min CSA. It flows slightly more as a bare head on the ‘bench, but then in reality are we driving around with no intake manifold? Well of course not. Bolt that back on and test it again, you see the flow figures are then equal.

It does give the ready made advantage of taking large diameter throttle bodies for high power/RPM use though, but then a 2.0 or 2.3 can be enlarged in just the same way in a few minutes.

Most 2.5 swap guys go to ridiculous lengths to blend their old 2.0 intake into the new bigger port. It’s a complete waste of time because the restriction is in the inlet manifold, not the entrance to the port. I’ve bolted all combos onto the flow bench and it doesn’t make a difference.

This where you get duped – most flowbench owners show you the CFM figures with a bare head only, never with everything bolted to it as it would be in real life.

It’s all about what power level you are at, as pointed out earlier the controlling parts of a port are the valve (throat) and lift, the Min CSA and the SSR so whilst they stay the same so will the flow – and it does.

The reason the 2.5 has a wider entry is to match the bigger diameter runners of course, these are used to slow down the velocity and weaken the intake pulses flattening the torque curve and making the engine less peaky. Exactly what you want in a van or heavy four door saloon of course which is why Mazda did it, it’s behaviour is very much like a diesel engine. Is this a bad thing? No, it’s just a different thing, you or I have to decide what kind of personality we want the engine to have and specify it accordingly – or just dump it in as is, no problem with doing that. 30ft/lbs more of torque and good economy? Yeah i’ll take some of that.

CNC porting

Having a CNC ported head does not mean the end result is any better whatsoever. A head had to be ported and developed by hand in the first place for the CNC machine to copy it. Porting is hard work end of story. If you can get a machine to do it for you then it’s going to save you time and also wear and tear on the human body. You can mess a head up repeatedly without having to leave your office with a CNC machine! Many do, lazy programming and the need for that all important CFM figure (sic) can remove many of the important port features without giving it a second thought. Copying a well modified head though with a CNC machine and yeah, now we’re talking.

So should you get your head ported?

It really depends on what power output you are looking for. Other engines from earlier eras had ports which didn’t have a fantastic shape to them so could be modified if you knew what you were doing. The modern car market is very competitive, you absolutely can not afford to leave power and economy on the table by putting a poor engine (with bad ports) in a car. The ports on Duratecs can be lightly modified to flow a little bit more whilst retaining the same valve sizes, but not a lot. You’re wasting your time and money trying to get it to flow much more whilst retaining the same valve size as Mazda got the ratios right from the start. Increase the port size and the next restriction is the valve throat, so next on the list is bigger valves, then you can port the heads a bit more aggressively.

If modified appropriately the 2.5, 2.3 and 2.0 heads flow enough for 300bhp and more without forced induction. As a rough rule I would say (N/A) that up to around 200bhp you’ll be ok on standard ports. 220 – 240 lightly modded ports, 250 onwards you’ll need bigger valves with associated porting. As you can see there is no cut off point, it’s just a gradual reduction the further you go and of course it’s highly dependent on other mods.

As far as forced induction is concerned then 500bhp is easily achievable without getting the grinder out, pushing it further than that and i’d be looking at bigger valves with some porting.


6 responses to “Technical: Duratec ports”

  1. I’ve always thought that intakes should be smooth, but not polished, while a polished finish on the exhaust reduced carbon buildup and helps long-term performance.

    So are you saying that short of ruining the port shape, smooth port walls will actually hurt flow and performance?

    • It depends on what you call smooth as you haven’t applied a method of measurement or scale. Certainly polishing an exhaust port is a waste of time and will lose power if applied to an inlet. A flow bench won’t care nor report whether the port is polished or not, but the engine certainly will.