It is 1994. Primus just launched the second generation of tds engines, which means turbo diesel sport. As a pioneer, the 1989 Astrona combined Diesel efficiency and reliability with the performance of a gasoline car. With the change from the Astrona to the Merit, Primus now introduced a six-cylinder replacing the old five-cylinder. The new benchmark in it´s class? Maybe, but there is some competition. And now that´s where you enter the chat. Some tried - as me - experiments with fake diesels, and some of them turn out shockingly realistic. One of the most diesel-focussed gamers is surely @Skygel14 .
I won´t take part, but compare your builds to mine just in a real car test. I know my car is good, but not impossible to beat.
Now let´s get to the RULES.
Car:
four door sedan, midsize to upper midsize (wheelbase 2,5 to 2,9 m)
model year 1988-1994, trim year 1994
five seats
no obvious nonsense like TRX cross ply tires, no mufflers or race parts.
Engine:
made out of cast iron or heavy aluminium (to simulate the heavier weight of a diesel engine, they are thicker)
turbocharger mandatory
Fuel: LPG
must behave like a diesel and have peak power between 3200 and 4000 rpm. Should be focussed on torque and early turbo spool, in the ideal case peak torque is at approx. 1500-1800rpm.
must pass WES5 (yes, thats lower than most countries have, but this is a Diesel and other regulations might be applicable)
max. cam profile: 15
max. spring stiffness: 15
not more than 50% balancing mass
no race parts
fuel map: max. 50
Techpool:
60 Million total cost combined (engine and trim)
Judging Criteria
HIGH:
Consumption (obviously, thats why you get a Diesel)
reliability (frequent travellers are the most obvious buyer circle, they hate being stranded)
comfort (did I already mention frequent travellers?)
MEDIUM:
drivability
sportiness and performance (the magazine wants to prove that Diesels can be fun, so no slug please, a hint: Look at the inspirations for acceptable values)
price (the company fleet manager should be willing to approve it)
design (because you are seen in it)
LOW:
service cost (most of them end up as fleet cars. Owning 100 money pits is a pain.
practicality (sedans are sedans, but some trunk space isn´t bad)
prestige (when you show up at 180 kph in the mirror the others shouldn´t bother moving over)
OOOOHHH I wanted to make a (series of) diesel challenge(s) at some point, but with my “willingness” to host things that could take ages to become reality Good to see that I’m not alone with such ideas, I very much like it.
I too experimented a lot with fake diesels, but since I rarely post cars outside of challenges, and those don’t accept “diesels”, there wasn’t much opportunity to see my experiments. Anyway, I have some conclusions to share, which might help with the rules here.
Fuel: LPG indeed tends to give the most realistic numbers overall. The efficiency goes up, there’s more “octane space” for increasing compression, the major drawback in terms of realistic numbers is the additional power/torque - although it’s not that much. Certainly much less than even more efficient LNG, which is why IMO it’s a better fit.
Torque (and power) and turbocharging: IMO there should be a distinction between CR and pre-CR diesels in that regard. CR very rarely, if ever, appeared without turbo, and it tends to make diesels much more capable in terms of torque and power - so if there’s any bonus torque from the fuel used, it won’t be noticeable, even compared to some real engine. However, for pre-CR diesels it can be hard to get realistic torque numbers, the harder the lower the boost is. Torque will be too high no matter what. That’s why I’ve decided to use a weird system in my “diesels” - those that don’t have common rail have their capacity reduced (in variant! to keep the weight) by ⅕ to ⅙ - it’s a range purely for ease of calculation, those are very similar values anyway. I’ve tested that against several non-common rail diesels in my replicas - turbocharged, naturally aspirated, 80s, 90s - and this tends to match their numbers.
Weight: IMO your solution is a pretty good one considering the era (were there any alu diesels back then?), but I think something more universal could be devised too, for future. The new blocks make this easier, with their light and heavy variants. I think a general rule of requiring only “heavy” blocks (incl. iron) could work, along with a ban on overboring - the logic being that you want to keep thicker walls in a diesel… and it also makes the engine heavier than the same capacity achieved with overboring.
Noise: well, diesels are a bit louder. That’s why I never use the first muffler in them. That doesn’t prevent them from being quiet, but makes it a bit harder to achieve/requiring more compromise.
Smoothness, or the lack of it: diesels vibrate more, and for a long time I had no good and universal idea how to emulate that. Less cylinders? Nonsense, that would mess up a lot of things. Forced 0 balance weight? I don’t like removing choices completely. But finally, I came up with something… a bit weird, but sorta working. Whenever an engine layout has balance shafts available, they’re banned. If it doesn’t, the harmonic damper is banned - as that’s the case in the smoother layouts, that isn’t much of an issue. However, due to the I6s length, it can be an issue for them, so they could be exempt from that, with the balance weight being limited instead. However… I think limiting balance weight to at most 50 (default) could be valid too, or in addition.
Revs: diesels rev low… but to a varied degree. The OM606 got its peak power at 4600 rpm! I’d have to check how my “diesels” are set up, but IIRC they all have 0 hardness valve springs - this naturally limits their ability to rev, and additionally increases efficiency. Also, I always make them idle low, but again, I don’t remember specific values. I think it was ⅕ of the redline?
Efficiency and emissions: AFAIK the diesel emission standards weren’t really that much different overall, and faking a diesel doesn’t seem to make the engine emit more. Except for one thing, that is also a total exception to my general idea of not forcing single values - the fuel map slider set to 0. Diesels burn lean and 0 map lowers the torque and somewhat increases the efficiency (although I’m not sure if it’s really optimal for that), and also makes the engine react much lazier, just like a diesel IMO.
I’ven’t experimented much in Automation, but as a mechanic who worked for CNH and a VW indy, among others, I’ve worked with real ones a fair bit. My $.02:
Real diesels vary. While typical automotive ones redline around 4500, and the party’s over by 4000 at most, I’ve seen powerbands as high as 6-7k in the case of a Benz OM606, or 5500 in the case of a Cummins 5.9. Internals are routinely capable of far higher RPM than the engine efficiently uses. A 1Z/AHU, for example, is supposedly fine hanging out at 8000 all day, just doesn’t make power up there.
Real diesels, especially one-wire turbodiesels, are quite easy and straightforward to tune for a lot more power and/or efficiency, typically at the cost of emissions. Datapoint: the 1983 Mitsubishi 4D55-powered pickup truck sold in the US, variously as a Mitsubishi Mighty Max or Dodge Ram 50, was rated for 28mpg highway for the 49-state 4wd version, and 24mpg for the California version, which differed only by having a more retarded injection pump specification. Tuning it by ear - advancing the timing until combustion noise took on a harder, “nailing” quality, then backing off a bit - with no further changes got our truck to do a consistent 32mpg highway and 27 or so combined, over several years.
Largely agreed with @Hshan, I propose that the best fake diesel look something like this:
** LPG, turbo, high compression
** very low cam, very low balance mass, very undersquare internals
** only one muffler unless it’s an especially modern/smooth engine, and depending on turbine size, potentially no mufflers at all
** fuel map zero, ignition fully retarded
** cast iron block, heavy forged internals, high quality bottom end
** fuel injection, perhaps with ITB for more realistic efficiency numbers, and to bring back a little responsiveness (diesels can be quite responsive as such, though typically with heavy, slow-revving flywheels). Rough equivalents: MFI = 1-wire injection pump, MPEFI = electronic injection pump, GDI = CRD
Here’s an example of something similar to a later Mercedes OM606 or earlier BMW M57 or M51. Note the powerband, reliability, and most of all the efficiency.
That’s a typical powerband. This next one is not, but still very much possible. Here’s an OM606 with a little work. 3.0 iron block straight-six with a DOHC-4 alu head, no VVT or VVL, old-school inline injection pump. 382kW (512hp) at about 4500-5600rpm. The engine can do more, still on stock internals; the next limit is the transmission.
I have never made a faux diesel in the current engine builder, but did lots in 4.1 (mostly Nissan inspired ones) and what you say is more or less the conclusions I did back then too.
Cam profile seems too generous. Without spamming quality, I can get a fake OM606 non-turbo to over 200PS. Fuel map restrictions seem unnecessary as well. As the 5.0 TDI V10 shows, diesels can be tuned to be just as gratuitously powerful and inefficient as gassers.
I’d suggest flat zero cam profile and balance weight, unlimited valve springs and fuel map. If someone wanted to build a 7k RPM firebreather, they still can, just not efficiently; such a motor would rely on boost, as they do IRL.
Alu blocks should be allowed but at a significant cost or power ceiling penalty.
Engine dimensions perhaps should be left alone, or perhaps mandated to be a certain bore and stroke over what’s actually used.
I’ve checked all the “diesels” I currently have in the game. The results are:
balance: basically only one had 50, that being a luxury segment V8 diesel, could work with 35 too; other engines usually had much lower values, one was around 35, so I fully agree with the 35 limit - still leaves some room to play with, but emulates diesel harshness (and I guess that’s more consistent than the balance shafts/harmonic damper ban with an exception…)
cam profile: all up to 20, but that 20 one was somewhat high performance for the era (100 hp/litre in 2008); most were even below 15. I’d argue for it being 20 if the challenge was set in a later year (say, 2008 mentioned before), but 15 sounds reasonable for 1994
spring stiffness: also all engines up to 20, the same situation as with the profile, and I have the same opinion
fuel map: every single on 0; while @moroza brings a valid point of engine tuning (which, in case of diesels, relies largely on pouring more fuel into them), I thing it’s irrelevant for stock, road car diesels - those are built for efficiency, without it they’re pretty much pointless; that’s why I’d argue for a limit even lower than 50 - not 0, as that would just remove a design choice, but for example 25?
One thing I forgot to mention before, that I also consistently use in every single of my “diesels” - yellow “performance lost to timing” warning. Why the heck would I want a yellow warning? Because that means I’ve traded performance for efficiency, which is, again, the point of using a diesel (well, less so with the modern, powerful CR diesels, but still). Also that helps keeping the performance in check (read: realistically low). Do with that idea what you wish, but I like that as a constraint since it doesn’t require any specific design choice, while making the engine more dieselish in a few ways.
Also I’d ask what’s your priority with the rule set - is it realistic effects, realistic design choices, ease of verification? Because different rules would fit those. For me realistic effects are always paramount, but it might be otherwise for you, and maybe less desirable in a challenge. Realistic numbers would require reducing capacity, but it’s a weird solution, and a bit cumbersome to use and check.
I partially retract the comment about cam profile. Attempts to simulate an OM606 turbo proved impossible without cam tuning (to about 18 IIRC). @Hshan asked some excellent questions at the end; pending answers to them, I’d say we take the latest proposed rules and just go for it, see what happens.
The yellow timing warning is hard to get with LNG.
I’ll add that “realistically low” performance is more due to emissions than some might realize. With relatively little work, VW 1.9 TDI has no trouble doing 150-200hp, OM606 250-300 (and upwards of 670 on stock internals), and let’s not forget the technical substance behind “Dieselgate”; it happened for a reason.
A low balancing mass could be a good compromise since that means that the revs have to be kept low for reliability which reflects IRL diesels to an extent.
But see, that’s with turbo, and hp is not the full picture. My point is not about limiting the maximum available performane, but rather the minimum. TBH with turbo, and especially with high boost, it’s not really that relevant. But the lower the boost, the more real and in-game diesels diverge. I remember trying to replicate the N/A OM606 and IIRC while I could get down to 136 hp, the torque didn’t wanna get lower than ~250 Nm at 3.0 litres, while the real one makes 210 Nm. Although, given that we’re supposed to make “sport” 1994 diesels here, I guess this might not be all that important.
As for the yellow warning - yeah, with LNG it would be very hard, but with LPG it’s no problem.
I would aim for realism in that aspect. I mean, the inspirations had between 90 (Passat, but there were soon variants with more power) to over 150 (E290 TD). I guess as automationverse seems to be always a bit above real life values, a power between 140 and 170 horsepower would be totally accurate.
This might be a silly question but should we just be using multi point fuel injection? With at least some diesels of this era having mechanical fuel delivery still I don’t know how well the mechanical injection in the game would carry over to that.
SPEFI might be considered equivalent to an inline vs. a distributor pump, but it’s not a neat analogy. None of these are, really.
For 88-94, MFI and EFI should both be ok. I’d argue for an MFI cost break but the experiment has enough variables already. As for GDI… the first CRD came out in 1995 (a medium-duty truck) and 1997 (normal passenger car), so if techpool can unlock it… I’d say ok to use it. What do others think?
Well, early CR engines weren’t yet that different in terms of their parameters, so I guess it could work. By default GDI unlocks in 2001, right? So with the standard +5 techpool that’s 1996, meaning CR would require additional investment, which sounds fair. Also, it’s worth noting that the first working prototype of common rail appeared as early as 1985, in a single IFA W50 (from East Germany!), but series production never occurred due to lack of funding. So even the maximally advanced 1986 date for CR unlock would have some realistic basis, although at that point it’s basically tech pool spam and cheesing I guess.
Good to see that I’m not alone with such ideas, I very much like it.
