Originally Posted by Jafro

Personally, I wouldn't try it. This kind of thread has popped up from time to time, and I always feel obligated to say something... I'd junk a 8-valve 1.5 liter. That's what I did with mine.

While it "could" be possible to build a D15B sleeper, some things you'll need to make it hold boost reliably will all have to be custom-fab. The valves suck, the manifolds are tiny, the rods are light and weak, the pistons are crap, the fasteners are junk, the crank has thin rod journals, and the wrist pins are dinky. The D-series block's floating-sleeve design is a recipe for disaster with boost. You can't do it without sleeves. Who makes a sleeve for a D15? You can buy at least 4 D15 short blocks for what the sleeve costs.

...so you need to replace the head, the manifolds, pistons, rods, crank, sleeve the block, use all new ARP chromoloy fasteners, dual valve springs and a different cam... ...then you might be able to hold boost... and maybe make about as much power as a B18 LS motor. The grinding, boring, honing, pressing, turning, decking, o-ringing, porting, sleeving... labor is probably going to be around $2500-$3000 because of all the custom work that will be required, and expect to pay around $2500-$3000 for parts depending on how much of it needs to be custom fabbed (crank, pistons, and sleeve). Headwork ain't cheap either. I'd just replace that head outright.

I'm not saying I think building a D15B wouldn't be cool. Hell, I did it with a D15B7, but at least mine was a 16-valve motor. The D15B8 is an 8 valve motor, and it's pathetic. You'd better have money to burn, because you'll probably end up with less motor than a stock B18 for more money spent.

If you want to go fast, there are other cheaper alternatives with just as good of a power/weight ratio. Like a Hyundai Excel hatchback with a $400 2.0L 4g63 NA swap. It's a bolt-in 140hp/140tq affair in stock NA fashion. That engine can hold 400hp on STOCK rods/pistons/crank if you boost it, and the "junkyard" turbo setup can be made from 100% factory parts with no custom fabbed anything. Maybe it's just the DSMer in me telling you this.

The machinist in me is saying... unless you want a museum piece that everyone questions the logic of, I'd scrap that idea. Yeah it would be cool, different, but why all that trouble and cost? If you installed that engine and tranny on a bar stool, then it would be really sweet, but put something better in your civic than a D15B8. There's no question your old motor was a great motor with 277k miles on it, but you got lucky. My D15B8 only went 104k before it lost compression. If you do anything, just put a DX head on that sucker, and call it a day. It's worth 35hp at the crank, and an extra 2000 RPMs to red-line with a DX ECU. Mine makes as much tq as hp which is odd for a Honda.

Edit: There's a full write-up of D15 guts in my sig. The rods and pistons of a CX/DX are the same. The pistons are even the same... the 8v pistons still have 4 fly-cuts on the top for a 16 valve head... go figure. The head casting is the same between the two. The valves are identical. The 16 valve just has another 2 valve ports bored through the exact same casting. The intake manifolds, exhaust manifolds, throttle bodies, o2 sensors, ECU's, injectors are very different, so if you can just get the top-half of a D15B7 + the ECU, there's your swap. Don't go higher than 9.5:1 on your pistons, and for the love of God, don't boost or spray it. Check my dyno sheet, my motor's 9:1 with stock internals and a 1-angle valve job. Those figures are hp/tq at the wheels, and that's about what a stock 95 si hatch does. Not bad. Use a different gearbox if you autocross. I kept the CX gearbox, and I only have to shift once to reach 80 mph. 3000 RPM's in 5th gear is 92 mph. It's not a transmission for racing, but I still get over 40mpg on the highway.