I've noticed that engines are rarely exactly 1.5 liters or exactly 4.0 liters, and I'm okay with that. It makes sense that engines would have some arbitrarily ideal size for specific applications, so I would expect exact engine displacement to be all over the place.
But it doesn't usually seem to be. Nissan's VQ35DE is 3,498cc, Toyota's 4U-GSE is 1,998cc, and BMW's N52B30 is 2,996cc.
Why aren't they just 3,500 cc, 2,000 cc, and 3,000 cc respectively?
There are regional taxation and tariff issues associated with displacement
Manufacturers will intentionally keep their displacement just under a limit that may increase their local/regional taxation as well import/export tariff regulations in accordance with trade agreements, etc.
It's easy enough to make an engine exactly 3 Liters. Math is exact if you want it to be.
Manufacturers will intentionally 'fudge' the numbers and round up for advertisement/marketing. 2 CC is a nominal fudge and a barely measurable performance increase and IMO not a misrepresentation.
Until relatively recently road tax (in the UK at least and presumably in many other countries as well) was based on engine displacement. As this worked in bands and engine which was slightly less than 2 litres would be a lot cheaper to tax than one which was slightly more. So nominal engine size made a big difference to the cost of ownership of a vehicle.
Equally, engine size is one of the headline numbers which determines the desirability of a vehicle ie a 2l engine sounds a lot better than a 1.5l engine on paper. So from a marketing perspective it makes sense to set your engine sizes close to the respective tax brackets.
However when you design the basic concept for a particular engine the capacity will be one of the first things you decide so you want to leave a bit of margin, If would be unfortunate if you spent a year designing the block and then a small tweak to the cylinder or crankshaft tip it over into the wrong category for no measurable gain in performance.
After all the difference between 1998cc and 2000cc is only 0.1% and you really really don't want to have to redesign and retool your engine plant to make everything 0.1% smaller if the government takes a different view to you on what is an acceptable tolerance, especially as an engine design may last for decades.
Also displacement is the sum of a lot of different parts, even disposables like spark plugs and head gaskets can make a measurable difference so even with pretty tight manufacturing tolerances you will see some measurable variation from one nominally identical engine to another.
From a more historical view I will put out the idea of "Dimensional Lumber"
Dimensional lumber is where a 2 X 4 is not actually 2 inches by 4 inches when you go to use it.
It varies by wood type and processing methods.
With that said way back in the cubic inches displacements (302, 327, 460 and such) the values were not as posted. 1940's 50's and 60's Chevy 216 = 216.48, 235 = 235.49, and 261 = 260.9. This was not for taxation.
If a manufacture wanted to make a finished engine with a precise displacement of a fixed value then they could start with the finished value and design backwards calculating in all the design and machining variables to find the starting values. But that is not the process.
But from the drawing board to the working model adjustments are made. Piston domes change, bores are changed for ring design, valve placement is adjusted, stroke is varied by crank and piston rod mods and machining, deck height changes, gasket types and even thickness adjusted. Some of which factor into displacement.
In addition factors relating to expansion and contraction of engine parts, allowance for oil lubrication change dimensions.
In the end you're close to a value, but it is not a perfect cylinder with perfect matched flat tops.
Last some of the odd values are based on parts exchange or squeezing out that extra bit of displacement. If a manufacturer can build one block but make two different engine sizes into it (one for economy, one for power) so much the better. Take engine "A" adjust stroke and make engine "B", (for instance 350 to 383 chevy) push the stroke to the maximum and you have a odd number of CID. Same goes for piston upsizing or down sizing.
Slightly different to the other answers but with motorcycles, there are often laws that restrict the usage (with your license) by the capacity in CC.
Sometimes these laws state that the size of the engine can not be greater than 125cc for example.
However, in other countries the same law may state that the engine size must be less than 125cc
Therefore, manufacturers like to maximize their sales by producing 124cc engines and marketing them as 125cc motorcycles
The size you are referring to is a measurement of volume of a cylinder.
V = pi * radius^2 * height
When you are multiplying by pi, it is hard to get an even number. Also, they care more about how it works than making the numbers nice and even.
In more an engine calculation Displacement = pi * (bore / 2)^2 * stroke * #cylinders
I think a better approach of looking at this is: Why, when the engine displaces 2,996cc's, do they call it a 3L engine?
While displacement of the engine is primarily what we're looking at, it all boils down to marketing. As @rpmerf stated, it's hard to get an even number. From an engineering standpoint, why even worry about it. It won't make that big of a deal in performance and would cost more money (in most cases) to come up with the alternative. So it boils down to it being a marketing standpoint. Saying an engine displaces 3L is a lot sexier than saying it displaces 2,996cc's. It will ultimately help sell cars.
This plays out particularly well when you consider what Ford did for decades with their 5.0L designation on the venerable Windsor small block 302ci engine. It actually only displaced 4,942cc, which if rounded correctly, should have been called a "4.9" and not a "5.0". 5.0 is a lot sexier than 4.9 from a performance standpoint. It wasn't until the new Coyote engine did the "5.0" actually become correct @ 4951cc's (rounding correctly this time, of course).
Rounding up is common and considered good marketing. Governments may base vehicle taxes, by whatever name, on round numbered values, 100cc steps, 850cc and below, 3.0 liters and above, etc. Should you race your vehicle, racing classes are almost always set by round number displacements.
But this is the best answer, because pi's irrationality (it goes as many digits as you care to enumerate) and non-roundness make a 3.000 litre or 1.600 or any other round displacement difficult to hit within .5 CC of a target value. (0.5 / 1500 = 1/3000 = 0.000333...) Round number displacements mean making piston bores, piston diameters and/or crankshaft strokes with 5 digit precision, non-round, values. If you're making zillions of something, any dimension is as good as any other, although that +/-0.0001 part tolerance is going to cost real money.
As the orders fall from millions to thousands, getting suppliers to make (3.14159 X 25.4)mm parts, +/-0.01% gets expensive too. Training people setting up the machines, ordering stock material, is easier, if you're making 76.2mm pistons, compared to 79.796mm. US built engines used to have 3.000 or 4.000 inch pistons, +/- more than 0.001". Far easier to make an 1149cc motor and sell it as a "1200". That's a 0.5% difference. No big thing. Unless you're trying to hit 1200 +/ 0.5, where you need 1/2400, 0.0417%.
I would guess that the reason is a combination of the desired performance, "real world" factors like taxation, and practicalities – what size pistons do you already have, how many parts can you gang at one time on your machines, what does your shop floor transport system look like, etc. Then you give it a designation that makes the marketing people happy and start building… Keeping in mind that while we think "4.951 x 10^3 mL" might look awesome on the back of a Mustang, somebody in sales might dread having to explain it.
I don't think the math/rounding argument explains it at all, with just close tolerances (±0.01 mm) you could have displacement errors well under a 0.1 cc. I think the train of thought goes something like "how close can we get to X liters (or cubic inches) without going over (taxes/insurance) and without having to redesign everything – and keeping in mind this big pile of other constraints."
Clearly there is an engineering component that cares less for pretty round numbers, this is evident in motor racing where the displacements are all over the place. HOWEVER, consider the tax consequences mentioned earlier for production vehicles. Let's just take the example of Turkey. From an article by Peter Mock for the ICCT:
"New passenger cars in Turkey are subject to the general value added tax (VAT), which is 18% and applies to all goods. In addition, a special purchase tax is levied, in Turkish called Motorlu Taşıt Araçlarına İlişkin Özel Tüketim Vergisi (ÖTV). In addition to cars, this ÖTV is also levied on tobacco, alcohol and various other items that are considered luxury products. The amount of ÖTV to be collected depends on the engine displacement of a vehicle and ranges from 45% to 145% of the vehicle’s base price. An important tax threshold is at 1.6L engine size. Above this threshold, the taxation level doubles from 45% to 90%. Another threshold is at 2.0L, above which the taxation level increases to 145%. The impact of this tax design is quite dramatic: For a new car worth 20,000 Euros the sales tax is 9,000 Euros if it has an engine displacement of 1.6L or below, but 18,000 Euros if it is 1.7L or above. No wonder that 95% of new cars in Turkey have an engine displacement of 1.6L or below."
Tell me that does not make a difference when you're selling hundreds of thousands of "2 liter" Honda Civics annually, each of which actually displace 1996cc.
