I am glad that has been acknowledged.

just to qualify i don't know what the sizes will be and barbz as yet hasn't mentioned that... he has found another issue that is causing back pressure..whether he will want to do anything about the size of them or even use the manifold at all i don't yet know but i will post what happens and why .

Yes, it is right that a lot of people, especially in USA started to use long runners - but why...
They started to use very big turbochargers and did not find any place under the bonnet to install them – they had to use long runners to find a place to install the turbo- best example are Subarus etc - then some people might misunderstood this and thought, this would be newest technology…
I really don’t know the reason for very long runners.
A turbocharger needs one thing – kinetic energy coming from pressurized air. When do we get most of it – directly after the valves of the engine, if the exhaust fumes are REALLY HOT. Hot air has a very big volume -> this results in very big pressure (Otto cycle), which the turbocharger can use.
Examples: High power diesel engines like the new MT 892 HPD (tank diesel which does not need any emission control) use post-injection to get more exhaust temperature – this results into a better response of the turbocharger and more top end power (diesel engines have a exhaust temperature of about 800°C max…)
The same procedure does Anti Lag – extremely late injection of fuel, which exaggerated explodes within manifold and creates tons of heat…
Therefore the charger starts to spool in low engine revs…
But what do bad engine tuners… build 1m runners to lose more than 100°C within this long way towards the turbocharger…bad job.
What is the reason to use a certain length of the runners – a single runner (at turbocharged cars) should avoid interactions between the single cylinders.
Problem here: The cylinder being within power stroke should be unaffected by the other cylinders.
Because of valve overlap it can occur that certain cylinders cannot get rid of their own hot exhaust fumes because other cylinders (having been in power stroke one cycle before) try to compress their own exhaust fumes into the opening valve port of the actual working cylinder.
Best example is cylinder number 4 of of original Fiat Coupe 5cylinder turbo manifold. This primary simply is too short – therefore a melting of pistons normally starts here.
Therefore there has to be done a calculation – how long must be the primary to avoid this interaction – but not too long to avoid too much power loss because of lower exhaust temps etc.
Newer turbo engines go another way – they simply separate cylinders with different power strokes and use twin scroll turbos. No there is no interaction any more between different cylinders and now very short primaries can be used to get most of kinetic energy out of the exhaust…
Unfortunately a 5 cylinder cannot use this technology
Juergen

Hi Juergan,

The long runners used by some aren't just due to having no choice due to turbo size and location, there is real effort to harness the pulse energy that the engine produces and for this reason the longer pipes are needed. To use a short runner manifold makes the potential use of pulses impossible other than at very low rpm which really there is no point in. If anything there is a lot of negative work being done at higher (more useful for a powerful engine build) by the pulses when the manifold runner is very short, reversion is high and the pulses are pushing exhaust gases back into the cylinder even harder than they would be. A longer runner is far more able to resist cross contamination from adjacent runners and when the rpm is higher it is possible to assist the exhaust gases in their efforts to evacuate the cylinder. This higher rpm is where the ambitious engine lives and dies, every help it can get is good. The short runner just backs up and temps go through the roof and misery is close behind. A 1m runner is a bit over the top, about 2/3 of that length is relevant and isn't that long really. Too long for a standard position turbo as the runners end up in a right tangle but with a turbo thats positioned differently the runners are potentially able to be routed in a simple free flowing manner.

The point of highest kinetic energy is just underneath the valve seat (smallest cross section the exhaust gases will see), after the gases have been accelerated out of the cylinder the port expands towards the exit off the head, this is how exhaust ports are in all (for the sake of discussion) cylinder heads and this expansion is what assists in MASS flow. This is what the turbo cares about, MASS flow. Move more gas through a turbo and it will work harder. So having expanded the port towards the exit in a carefully ported cylinder head there is no sensible reasoning to then attatch a runner with a smaller inner diameter than the port exit. A 36 mm id runner has about 20% less area than the exit, this can only restrict mass flow when compared to a runner that is close to the size off the head. I've searched long and hard and no single person has managed to explain why a restriction to the gas flow is an advantage, maybe you can?

The exhaust gases are under pressure, even a very well built turbo engine will be lucky to get to a 1:1 ratio of intake pressure/exhaust pressure. This pressure is what keeps the gases moving through the turbo, all they are interested in is getting to atmo. To increase the exhuast pressure doesn't increase the mass flow, just the same as increased boost on the intake side doesn't increase mass flow. The gas momentum needs maintaining (the point under the valve seat gets everything moving plenty fast) rather than hindered.

Regards,

Nik

Jeurgen, I'm totaly with Nik and others on this whole ongoing primary dia topic.

Your considering heat as too big a factor, yes it helps to spool a turbo, but on it's own heat is useless. Pressure ratio is what drives the turbine, again - anti lag does not spool the turbo in low engine revs by the heat from an explosion in the runners, the pressure created by that explosion is what spins the turbo.

In fact imo loosing some heat in the exhaust gases is a good thing, a little less heat expected means you can run a bit leaner mixture this in turn produces more power, and improves BSFC.

Another point I'd like to make is when you talk about valve overlap and exhaust gas interference across cylinders, I don't think you fully understand the concept of interference and cylinder scavenging. This happens at the collector and is worked out not on the relativly slow moving gasess, but by sonic sound pulses which can CARRY gases with them to a degree. This is the fundamental reason for a designed length of runner, the length is calculated from the varieing sonic speed of these resonant pulses which reflect off a change of cross section, wheter that is seen in the collector it's self or in the pipes at a purposely designed stepped section. Another problem (and the point I want to make) is that with a small diameter pipe these sound waves are actually amplified in there intensity, and can be deadly to an engine if they are not 'timed' to coincide with vavle events by way of a certain length of travel ie; runner length. I don't see any reason why a 5 cylinder cannot use these pulses effectivly, but it would take some serious simulation.

Twin scroll uses this to good effect on OE setups because the cast, short runner manifold can be 'packaged' into a production engine bay, there is no other reason for using TS turbos other than that. We have the abillity to change our runner length and alter scavenging how we please.

On another note I don't understand the obsession with 'spool up' some people have, It's a turbo car at the end of the day if you want response build a naturly aspirated version. You can make more power at the other end of the scale anyway ie; far better to rev to the moon :-)

I understand the physics of a otto cycle fully, but it seems that you don't...
A turbocharger does not live from Mass flow, only - this is completely wrong! It needs a difference in pressure - means a difference in temperature.
If it is this way (living form mass flow only) - please install the turbocharger at the end of your exhaust. There you still will have the same mass flow like directly after the valves...
Only difference at this location is a loss of temperature!
Pressure and temperature is COMMON. You won't get pressure without temperature...
Please have a look to thermodynamics!
If you are right, then all OEM are idiots:
look here (has been one of the engines of the year)
http://www.bmwblog.com/2009/11/12/diet-turbos-how-low-lag-is-no-lag/
Information: BMW had very very big problems to get the turbos into this location - but they wanted to use all of the kinetic energy to get a good engine...
Thus for example they had to construct a special wire harness, which can do the heat, the ECUS got their own cooler etc. etc...
All modern engines have extremely short runners meanwhile. If you show me ONE modern engine, which has been engine of the year and which has long runners I will believe you :-)
Juergen

Well I'm afraid you contradict yourself because paragraph two of the link you just posted reads "There is tremendous energy in the flow of expelled exhaust gas leaving the engine.

This is where it's at, the turbine converts presure to velocity. end ov.

I can't help you Juergen, when you take your inspiration from OEM's and engine of the year awards. I take my inspiration from winning drag teams and real obsesive petrol head types. Watch this clip in full, proper tuning and performance in a car with 5 seats and petrol bought at the pump.

Look at the primary pipes, and notice how lairy the car is, not much "lag" as you call it to worry about there either. Come back and tell me those headers are too long or big in diameter.

http://www.youtube.com/watch?v=skly2gsdbyU

Coool - a "street car" which uses a anti lag system (listen to the engine sound) to get power in low revs...
I assume that this car will make peak power in 8000-9000s and before 5000rpm there will be absolutely no power without antilag (which will destroy the turbocharger definitely after some time)...
Do we talk about street cars or do we talk about 1/4 mile racers which won't survive one minute German Autobahn!
OK Tricky - I understand you
OEM's are idiots and ktm450exc want's to build a 9s Fiat Coupe.
Therefore I am now going to share your opinion:
Rob - your 36ID- manifold is not big enough to reach your 600+ HP.
My car, producing 565whp is not true, because my 34mm ID manifold cannot do this amount of flow...
No comments any more from my side to this topic.
Ps: I have edited my statement to mass flow and its physics behind it for you to understand!

Juergen

[quote=tricky]This is where it's at, the turbine converts presure to velocity. end ov.[quote]


Not sure you have that fundamental right. The turbine converts the kinetic energy in the exhaust into work done in rotating the compressor. The formula is: kinetic energy in the exhaust is equal to half the mass of the exhaust flow x the velocity of the exhaust flow squared, or Energy = 0.5 * m * v^2 - so you can deduce that gas speed is critically important.

For a given BHP, the mass of exhaust is fairly constant and accurately predictable within a few percentage points.

The speed of the exhaust flow depends on the volume of exhaust gas being flowed and the cross sectional area it is passing through – e.g. a valve seat area, a primary pipe or a collector/turbine entry.

The volume of the exhaust gas is dependant on the density of the exhaust gasses.

The density of the exhaust gasses is dependant on the temperature of the exhaust gasses.

A quick calculation shows that with a 2.5 litre engine producing 500BHP and a turbine entry temperature of 950C, about 73 HP of energy is contained in the exhaust gas at the entry to a T3 turbine housing. That falls to just 37 HP at 600C turbine inlet temperature – lower because inlet velocity will have fallen so much. In fact with only 37HP available, the turbo would not be able to flow enough air for the engine to produce 500BHP.

On the subject of choking, this only occurs as gas speeds reach Mach 1. The speed of sound is dependant on gas temperature, so in exhaust gasses with a mean temperature of, say, 850C the speed of sound is 690 metres per second. If the engine is operating with WOT, when the exhaust valve opens initially it will be operating in "choked flow" as the gas speed will be at Mach 1 at the valve seat.

At maximum flow, pressure in the manifold will be close or equal to the boost pressure the turbo is producing from the compressor side, despite the fact that 30%-40% of the exhaust gas is bypassing the turbine and exiting through the wastegate. This pressure is therefore substantially above the 1 of 2 psi extra any back pressure that smaller primaries might add to the system. As the pressure inside the cylinder will be far higher than the exhaust manifold pressure during the exhaust period, the differential in pressure across the exhaust valve will ensure the cylinder is emptied of all or most exhaust gasses.

Hi Juergen,

I've just had a read through some of the posts since this morning and to say there are opinions flying about is fair! I didn't say a turbo works on ONLY mass flow but rather it's this that it cares about. I'll have something to eat and try and write again afterwards in an effort to address some of the various opinions. I will say at this point though, we're discussing an engine that produces far more power for it's capacity than anything a mainstream manufacturer supplies to the public, the comparison is not relevant I'm afraid. I'll bet though that if you measured the manifolds off production cars you'd not find any that have a smaller cross sectional area than the port exit. Usually the problem is the exit is too big not too small. Show me a manifold that is a smaller size area than the exhaust port and I might start to sway a little :-) We'll not find a long runner manifold I'd agree but thats due to totally different design goals that they have (oe manufacturers) when compared with fella's that like to build high power cars such as we are talking about.

Regards

Nik

Very good explanation, Thank you very much.
Only additional comment from my site:
Chocking of exhaust flow starts at 0,6x Mach 1 (about 200m/s@~20°C) - this was the formula, I learned...
At higher temperatures Mach 1 speed is much higher, as you stated, it is not constant...


Ciao
Juergen

This is commonly quoted value and I am in no position to argue with it, though if you do some Internet research you will find others who are qualified and have argued against it extensively. It seems that the value "Mach 0.6" varies according to how the flow is affected by local turbulence - i.e. in a smooth straight pipe there is very little to cause turbulence, but splitting air over a wing section causes more flow separation at higher velocities and therefore more turbulence. What I will say though is that beginning to choke and actually choked are not the same thing; beginning to choke means that the rate of flow is not increasing in direct proportion to velocity, but flow rate will still increase with increased velocity up to the point of sonic flow (when it will remain constant) - it is just less efficient.

I have some "ball park" figures that I work with that illustrate the problem nicely. A well designed inlet valve and valve seat will flow efficiently up to Mach 0.333 but a well-designed inlet port will flow up to about Mach 0.533 before efficiency suffers. Neither of these situations can be compared to a constant diameter, smooth bore pipe, as both clearly require the air to make some radical direction changes around obstructions - the valve guide/stem, the SSR, the valve head and seat. The valve/seat combination is obviously more restrictive than the inlet port, but it is well known that a poppet valve has a discharge coefficient significantly less than 1.0.

Like it or not, flow in the exhaust will (at exhaust valve opening) be sonic - the pressure differential between the gasses in the cylinder and the pressure in the manifold of 6+ bar, and the must smaller effective area of the valve or exhaust port (whichever is the most restrictive) compared to cylinder area assures that - and there is nothing that can be done by the engine designer or us to avoid it.

However, going back to our 2.5 litre engine with 600BHP, 36mm primaries and EGT of 850C, the mean exhaust gas velocity in the primaries will be 283 metres per second, which at 850C is only Mach 0.41.

I agree with needing a pressure difference, why you'd increase the pressure in the manifold (where the gases are trying to get through) by using a runner thats smaller than the port I don't know. Thats just hindering ability to move from high pressure to low.
With regard to putting the turbo at the end of the exhaust we've been through this before Juergen and you didn't offer reasoning then either, please refer back to the post where I showed you a very fast vehicle that did have the turbo a long way from the engine, relevant pictures at the bottom of that page.
http://www.fiatcoupeclub.org/forum/ubbthreads.php?ubb=showflat&Main=90133&Number=1085252#Post1085252
I know about the temps involved with turbo charged cars, if you want to keep them sky high I can only wish you well. I consider the volume of exhaust as massive and easily capable of spooling turbos even when silly people put the turbo at the end of long runners.
The BMW link, I'm afraid I switch off when I read "During the ‘Exaust Phase’ of a four cycle engine, the exhaust valves open and the piston forces the combusted fuel/air mixture out of the cylinder with a plunging movement." It's an article written by someone who knows very little. It's hardly an engine to compare with what we are discussing though is it, whats it got, 50 bhp per cylinder? Thats not quite the same as 100 bhp per cylinder on your 5 cyl or 150 bhp on a real engine like a 16v 4 cyl is it
Instead of refering to production engines I'd advise seeing what people do with performance engines that break the rules you are living by. An example a friend of mine helped design and build is worth quoting you, this breaks the rules you say are impossible but it's very true and not rare, there plenty like it around this worls.
"One example is a 1.9-liter inline 4, max revs 8200 rpm, 25 psi boost max, reaches this at 3870 rpm in 3rd gear as per the log the customer showed me.. That little thing makes 5 psi at 1700 rpm - and, already by then it pulls well and I would not call that an "useless power range" - and close to 420 whp (actually hub hp). At 25 psi boost and 7700 rpm the EB was 22 psi."
That's using a GT35 based turbo, completly true and more than possible.

Regards,

Nik

Nik

I completely agree that it would be wrong to have primaries with a CSA less than the exhaust port exit at the head - no argument there at all. However, the debate was about whether 36mm primaries would choke the engine, which they won't. The port area was not mentioned, although the valve size was later in this thread. I would argue that the effective valve area is in fact the area at the diameter of the inner valve seat less the area of the valve stem. In that case two 38mm valves with 7mm stems have a flow area of about 900mm^2, which is equivalent to a pipe of about 34mm ID.

However, I agree as a rule of thumb with using valve curtain area as the basis for calculating primary area - but I would use it to say that there is no point in going to primaries larger than that area. In a N/A engine, without the massive back pressures which a turbine in the exhaust flow causes, I would also be looking to minimise back pressure, but in an extractor manifold that is usually achieved by ensuring a high velocity to cause low pressure in the collector/s to provide 'suction' at the collector end of the primaries.

The debate about long and short exhaust manifolds on turbo engines is an interesting one. Subaru, and of course Porsche before them with their single turbo engines, had no choice but to use long runners on their boxer engines – and they can obviously work very well. Short runner manifolds are able to utilise the pulse energy in the exhaust – perhaps better than long runners - but there is no opportunity for wave tuning. Pulse energy gives greatest benefit in spooling up the turbo when the mass flow is low. However, at higher rates of mass flow the pulse energy becomes less and less significant to the performance of the turbo. What does that mean in real life? Well, the way I see it is that an engine that is often used to accelerate from low through medium to high speed will make better use of pulse energy than an engine that is always operated at higher engine speeds – so a road or rally car would do better with a short manifold and a race engine would do better with a long manifold.

The other thing about pulses is that they get reflected back down the manifold and this can be detrimental to performance. In a four-cylinder engine there is always one cylinder on its induction stroke while another is on its exhaust stroke. Should a positive pressure pulse from the cylinder on it’s exhaust stroke be reflected back and arrive at the wrong time it will add to pumping losses on the cylinder on it’s exhaust stroke, but worse still impair the cylinder filling in the cylinder on it’s induction stroke. Very short runners should ensure that the positive pulse reaches the exhaust cylinder before it reaches BDC where it will have the least impact on pumping losses, but at this time the cylinder about to start its induction stroke will have both the exhaust and inlet valve open, so the positive pressure pulse may cause reversion and push some of the incoming charge back out of the cylinder.

The conventional way of dealing with this problem in a 4-1 manifold is to have runners long enough to delay the pulse, such that on its arrival back at the exhaust port on both cylinders the exhaust valves are closed and the pulse will cause no harm. In a 4-2-1 manifold, cylinders which can’t interfere with each other are paired – in most cases this is 1&4 and 2&3. Because these pairs don’t interfere with each other the pipes connecting them – i.e. the 4 into 2 part - can be much shorter (though the total length to the final collector of the 4-2-1 manifold will be much the same as a 4-1 manifold). The shorter manifold benefit of connecting non-interfering cylinders together in this way can be enjoyed on a turbo engine if a divided inlet turbine is used; pulses reflected off the turbine wheel can only go back to the two cylinders that are connected to the same inlet.

Looking at the BMW article mentioned above there is a picture of the F1 Brabham powered by the four cylinder BMW turbo engine. You will notice that the primaries are quite long – certainly much longer than on the V6 turbo cars of the same era. This is because the three cylinders connected to their own turbo on either side of the V do not interfere with each other, as their firing cycles are separated by 240 crank degrees rather than 180 crank degrees on an IL4. The Hart 4 cylinder turbo of the same era also ran long primaries. An educated guess is that the reason for this is to reduce the power losses that would come from reflected pulses interfering with induction, adding to pumping losses or both.

So on an engine that has interfering cylinders connected to the same, single entry turbo, longer runners would probably give more top end power. Designing one to do the job is a complicated business though. The maths is quite straight forward, but defining the task is not so easy.

The first complication is deciding at what engine speed you want it to ‘work best’ – because what will work at high RPM won’t work at low RPM when the time between valve events is longer. The second complication is at what exhaust temperature you want it to ‘work best’. This is because the speed of the pulses is sonic, and the speed of sound varies with temperature. So, higher revs = shorter runners, higher temperature = longer runners. As most petrol engines will have a fairly wide rev range, and due to varying loads widely ranging exhaust temperatures. You will also need to know exactly when the exhaust valve opens and closes. Deciding these parameters and ensuring the engine operates within them is key to designing a long runner system. A long manifold is, in my opinion, much harder to get ‘right’ and easy to get ‘wrong’.

I think this is the main reason, apart from packaging and perhaps better spool up, that short turbo manifolds are most commonly offered in the ‘aftermarket’ and that longer manifolds are really best suited to specialist applications like circuit racing, or forced on the aftermarket because of packaging in the case of Subaru!

I was tempted to quote again but thought better of it, it's a huge page now!

I'm sick of picking posts apart so will resist now, we could go on and on without end.

Just for interests sake though I've just been measuring a 20vt head and it's exhaust ports. I'll admit to not being 100% accurate as I didn't pour a port mold to measure but measuring carefully with some tools I have and crunching some sums I see how the unported head has an exit out of the exhaust port of just less than 1300m2, not the big gapping exit but the expanding section as it moves away from the seat. There is no doubt in my mind that a 36 ID runner does nothing other than restrict the very thing we'd like to be free flowing. Cross sectional area is where it's at, carries right through an engine. This was the discussions point wasn't it.

group5lancia, not critisim I enjoy your posts but just a few comments on what you have written.

36mm primarys won't choke the flow, but they will still form a convergeance when ideally we want a divergance or 'megaphone' effect to help with the PD. Valves flow very diferently remember depending on which direction flow is oriantated, an exhaust valves head doesn't pose a restriction like an inlet due to it's opposite flow direction in the same way gas is almost insensitive to exahust valve seat angles as opposed to inlet ones. Therefore the smallest controlling section is always the throat and never the curtain area, down the port then widens out to the gasket face as this trend needs to continue into the runner and therefore decide it's ID.
Thats my take on what the gases actually 'see'

Runner length is based on the needed powerband, max rpm and is largely related to cam timing as you have said. Just to throw a spanner in the works :-) I have had my own exhaust manifold designed by simulation recently by a specialist exhaust company using all the parameters from the engine (16v) with a powerband of 3000-8000 rpm and for my fairly modest cams a runner of 22" total was predicted, this is on a road car just for your interest.

I have my suspision what ID of tube you personaly would pick for lets say a nice 16v engine, but what size do you calculate is best ?

Just some other thoughts on 'lag' or 'spool up' people too often confuse this with boost threshold, ie' "My GT 35 is really laggy", no it's just a turbo that technicly is far to big for the kind of exhaust volumes your engine is capable of ever flowing.

The trouble with it all is that it's such a re occuring problem because the inlet side needs the mass flow to produce the cylinder pressure to produce exhaust volume to power the turbine wheel to turn the compressor to create mass flow, so we're back to square 1 again. It's a chicken and the egg problem, Mass flow from the compressor is where it starts NOT in the exhaust side imo altough saying that if you fiddle the exhaust cam timing a few degrees you can make better use of the cylinder pressure by way of early EVO which will pulse the turbine a little more at the expense of usefull work from the expanding gasses to drive the piston again another chicken and egg situation !

Lag is a transiant, if your charge air is'nt hanging around in the intercooler, there's not much else that can be done IMO. People seem to go to all kinds of strange lengths to irradicate it, like small primary tubes ! or ITB's or 2" pipework because it takes less time to fill ! But at the end of the day, just pick a turbo that matches your rev range and just wait for it !

Agreed about the better flow coefficient of an exhaust valve compared to an inlet valve. What I have hi-lighted in your post is what I was saying above about the 28mm valves - if you calculate from the curtain area you would end up with primaries which are larger than necessary.

Curtain area doesn't feature in the sum to decide the ports throat, it's the valve diameter that determines sizes throughout the port which in turn carries through into manifolding.

OK, but I believe calculating the primary diameter from the valve area based on the overall valve head diameter is wrong, because the flow at the valve is not determined by the size of the valve head, but the size of the hole the exhaust must pass through! i.e. the valve throat area minus the valve stem area. Anyway, it works for me!

Oh yes, I think we all sing from the same song sheet in this area then.

How do you calculate valve curtain area group5lancia? or do you have a simulation program that does it for you?

If you calculate it physically how do you cater for cam lobe profile varience? also, what circumference do you decide on as being the curtain area?

I calculate peak and mean - both from published cam duration and clearance info - or if I have mapped a cam, directly from the profile.

However, in my last post I was agreeing with you that the curtain area is the wrong thing to use when calculating primary area/diameter - it was a slip on my part. But I don't agree with using valve head diameter/area either as it doesn't represent the flow area - I gave my reasons above.

Now this is getting even more interesting, could you potentially use a combination of valve throat, even say minus the stem dia and the max valve lift to calculate curtain ? And is it of any use to us anyway ?

- Not moving over to the dark side are you Nik ;-)

not being funny guys but this is clogging up a post if you want to talk flow and valve design etc etc then start a thread about it.

I be following it anyway i just think it is unfair on ktm's thread

Can Rob not chase us if it bothers him or are you his keeper? It's a bit late to tidy it all up.

@ group5lancia
"i.e. the valve throat area minus the valve stem area. Anyway, it works for me!"

I've read it more clearly now, just looking at a phone before. The reason I refer to valve diameter is that it is part of what determines the throat size and inturn the runner size as you said.

@tricky
"Not moving over to the dark side are you Nik ;-)"

Not really but the 20vt is a great engine, I look through the classified section often enough and would buy another as I miss the one I had a few years ago.

You should come over to evocorner, you could spend a year trying to orginise the 500,000 long threads and not touch the sides

Ye, I suppose it's alright in it's own right. I would'nt throw money at it though.

Nobody is his keeper, but you are filling up pages with unrelated waffle.

We're interested to see how Rob's car progresses, not how much you can write about something that interests only you and a couple of other people.