David Vizard
March 1, 2008
Photos By: David Vizard

Step By Step

View Photo Gallery
Mmfp_0803_01_z DSS_347_stroker_engine_build Mmfp_0803_02_z DSS_347_stroker_engine_build
A stock, late-model, 5.0 block was used in this build to keep down cost.
Mmfp_0803_04_z DSS_347_stroker_engine_build
The bottom of the bores were notched for clearance.
Mmfp_0803_03_z DSS_347_stroker_engine_build
The rotating assembly consists of a Scat neutral-balance crank, I-beam rods, and flat-top pistons.
Mmfp_0803_05_z DSS_347_stroker_engine_build
Heavy metal was used on the throws to balance the crank. After sliding all the caps into place, the caps were torqued.

Almost every year, I get the opportunity to race in Trinidad or Grenada. It could be a road race or a drag event-either way, my crew chief of choice is Mervyn Bonnett simply because of his race-winning abilities. But crewing and racing in the Caribbean involves a lot more than you might think. With staggering import duties and the difficulty of getting spare parts, a crew chief's job becomes one of understanding the nuances of import and export and being able to produce winning performance both reliably and cost effectively.

A few months back, I got a call from Mervyn telling me one of his car owners wanted a "track day" small-block Ford for a car so he can build seat time. This meant the engine had to be reliable and have a power curve that emulated an all-out race engine so as to give realistic shifting practice. All the pointers aimed towards a 331-inch combination that could run on pump gas. Why a 331 rather than a 347? Because a 331 can turn more rpm with less block loads.

As for power, our goal was a target of 525 horses. This was more than achievable with a 331 and would give the block a better chance of surviving lap after lap of punishment.

Having built a couple of 5.0 Fords with D.S.S. parts, Mervyn wanted to stick with success since both engines made reliable power. The modus operandi here is that Mervyn comes to the United States and buys the engine as a kit of parts. He then assembles the engine, dyno-tests it, then strips it back down for shipping as used parts. The teardown also allows him to inspect everything to be sure all is well internally before he goes back to the islands. A failure when the engine is out of the country is a major deal since the duty tax on new spare parts going to Trinidad or Grenada is about 125 percent. Well, all this looked like a good opportunity to see what D.S.S. could produce for a track day engine.

The Block
The 5.0 late-model block from D.S.S. came with all critical surfaces machined. This included boring and deck-plate honing, decking, and align-honing the mains. The quality of the block prep was all that we have come to expect of D.S.S. Since blocks are shipped from the company virtually surgically packed, they're ready to take the rotating assembly with little more than a cursory wipe-over of the locating surfaces.

The Rotating Assembly
The rotating assembly decided upon was built around one of Scat's economically priced aero counterweighted, internally balanced 3-1/4-inch-stroke forged cranks. As you can see from the nearby photo, the crank took three pieces of heavy metal in both ends to achieve internal balance. The bottom line is that, in the projected power range of this engine, the crank will be nearly indestructible.

This Scat crank was paired with a set of D.S.S.' flat-top lightweight pistons and pins, which were to be mounted on D.S.S. lightweight billet I-beam race rods. Total Seal rings were chosen to fit the pistons' 1.5mm/1.5mm/3mm grooves. Using high-dollar micrometers both inside and out, the bearing clearances were checked. For the mains, the spread was from 0.0024 inch to 0.0027 inch, which was plenty close enough to our target 0.0025 figure. On the rods, the target clearance was 0.002 inch, and by selective assembly, we arrived at clearances from 0.0019 to 0.0022 inch-again, close enough for all practical purposes.

Piston and Rod Install
To get the narrow Total Seal rings to transition into the bores without damage required a tapered ring installer and a certain amount of dexterity. After all the rods were in, the bottom end fasteners (mains and rod big ends) were given a final torquing. At this point, a provisional turning torque test was done to ascertain that the assembly turned freely. With the rear main seal installed, this assembly should take no more than about 20 lb-ft to turn it. Ours turned at less than 13 lb-ft. The next job was to install the stud girdle onto the extended ARP mains studs.

Step By Step

View Photo Gallery
Mmfp_0803_06_z DSS_347_stroker_engine_build
Mervyn Bonnett carefully lowered a piston and rod assembly into the block.
Mmfp_0803_07_z DSS_347_stroker_engine_build
Next, the main girdle went into place.
Mmfp_0803_08_z DSS_347_stroker_engine_build
The timing cover and oil pan went onto the small-block next.
Mmfp_0803_09_z DSS_347_stroker_engine_build
The windage tray and oil pump were next to go on.
Mmfp_0803_10_z DSS_347_stroker_engine_build
A degree wheel was used to properly time the camshaft.

Main Girdle and Pump Install
At this point, the main girdle was installed. Once in place, the driveshaft and oil pump were positioned and, together with the pickup, bolted down.

The next item to be positioned was the windage plate. This design seems exclusive to D.S.S. We have not tried a before-and-after test on this windage/oil scraper plate, but since all of our D.S.S. motors have performed well, we have no reason to question its function. That's a test for a future build maybe.

Once in place, the oil pickup/pan floor clearance was checked. This should be about 31/48 inch. Usually only minor adjustments are needed to achieve this, and such adjustments can be done with a rubber mallet.

Cam Selection And Timing
The cam used in our 331 was a solid Hi-Tech 0.420-inch lift roller grind from Comp Cams. This featured an intake lobe delivering 266 degrees duration at 0.050 inch lift and 302 at 0.020, with 0.652 net lift with the 1.6:1 rockers we intended to use. On the exhaust side was a lobe with 272 degrees at 0.050, and 308 at 0.020 and 0.652 net lift. The lobes were ground on a 108 LCA and, to get the best power curve, the cam needed to be installed with the intake centerline on 104 degrees.

By any standards, this is a big cam, and there was a little concern that it may be too much for the job (i.e., make too much power at too high an rpm). But because the ramps on these profiles are designed to be dynamically easy on the valvetrain, we left that part of the engine spec as is. If you're looking for a durable, power-producing cam, these Hi-Tech 0.420 roller profiles from Comp are well worth considering.

To make sure the cam could be timed in "right on," Comp's adjustable timing set PN 8138 was used. Once the cam was correctly timed in, the front cover was installed, followed by the pan. A one piece Fel-Pro pan gasket was used here as it is a reusable item, allowing the pan to be pulled and replaced without the need for a new gasket.

Damper, Turning Torque, and TDC
The next item to install was the crank damper. This has to be done with an approved style of installation tool. The hot damper and hammer method is absolutely not approved. If you have to install a damper, Auto Zone has a loaner tools program, which refunds 100 percent of the deposit upon return of the tools.

At this stage, we can do a meaningful turning torque test as the long-block is now complete with all the seals, plus cam and timing gear. Adding the most recently installed parts upped the turning torque to 17 lb-ft. This is a great number for a short-block with otherwise-stock tension rings.

Our next move was to establish TDC for the purposes of knowing, and setting, ignition timing. This needs to be done before the heads go on. Doing it after is possible but usually not as accurate. A bore bridge dial stand as seen here is the preferred way to go, but the job can be done equally well using a regular magnetic base-just remember to set up the indicator on the piston centerline to minimize the effect of piston rock in the bore. The D.S.S. pointer shown here has to be one of the best we've seen for a small-block Ford.

The last job, before moving on to the cylinder head install, is to fit the water pump. Here you need to make sure that the pump's designed direction of rotation is in the right direction for the application. Engines before the use of serpentine belts require a forward rotation, and ones for use with serpentine belts are, for the most part, reverse rotation.

Cylinder Heads
Because of their taller-than-stock height, the lifters must be installed before the heads go on. To complement our Comp Cams roller cam, a set of Comp's regular race rollers, PN 838-16, were used.

Step By Step

View Photo Gallery
Mmfp_0803_12_z DSS_347_stroker_engine_build
Then the harmonic balancer was installed.
Mmfp_0803_13_z DSS_347_stroker_engine_build
The engine was brought to TDC, and a new timing pointer was bolted up.
Mmfp_0803_14_z DSS_347_stroker_engine_build
Taller-than-stock roller lifters were dropped in. This must be done before the heads are installed.
Mmfp_0803_15_z DSS_347_stroker_engine_build
With the head studs and head gaskets in place, the AFR heads were slipped on the block.
Mmfp_0803_16_z DSS_347_stroker_engine_build
Next, we set up the roller rockers.

With the rollers installed, the ARP studs were coated with a slow-drying, nonhardening thread sealer and installed finger tight. Although these studs have an Allen socket in the end, this is not so that you can tighten them down in the block-it's there so you can remove them easily at a later time. To achieve the correct preload, and load the block threads evenly, the studs must not be tightened into the block any more than finger tight. Be aware that if you intend to use a header adapter plate, there can be an interference problem with the regular ARP short outer studs. The fix is to use either shorter studs or go to ARP 12-point bolts.

With the studs installed, the head gaskets-Fel-Pro's latest multi-layer design-were positioned. These need to be laid on the block with the brass button upward, as shown nearby.

At this point, the heads went on. The ones chosen for the job were AFR's 205cc runner, CNC-ported, Outlaw street/strip heads. There has been a lot written about these heads and their near legendary performance. That they produce results is unquestionable, but after spending some time on the flow bench, there are a few facts to offer that may not have seen the light of day before, so check out the sidebar on our AFR heads.

Over the years, MM&FF has written a lot about the AFR small-block Ford heads-all of it pretty much glowing testimonials of their excellent performance capability. Where does this all come from? Well, some testing on a more exotic flow bench might reveal a few facts not mentioned in previous write-ups. First, here are the basic head specs so you know what we used.

The casting used is AFR's basic Ford offering but with CNC -machined ports and chambers. The intake port has 205 cc, which makes it slightly better suited to a 350-inch engine than our 331. The intake valve, at 2.080 inches diameter, is a little larger than is often used with a conventional Ford head configuration.

(Due to an oversight during the build of our 535 horsepower small block in the March 2008 issue (Stock-Block 535HP screamer, page 150), author David Vizard mistakenly reported that the heads used on the engine were the AFR 205 competition heads, when in fact the actual heads used in the testing were a pair of AFR 185 street heads which are actually emissions legal and far less expensive. While 535 HP is still a very respectable number, it's even more impressive coming from a set of street heads that retail for $1450! AFR is confident the much higher flowing, larger runner 205 cc race heads would have actually produced even stronger results in that application probably edging the final numbers a lot closer to 600 HP., Ed.)

On the exhaust side, the 75cc port is fed from a 1.6-inch-diameter valve, and the exit point of the port itself is raised 1/8 inch. This small amount should not affect header fit, but it does give the head designer the opportunity of increasing the short-side turn for better flow. The head casting shows its heritage in that the exhaust port shown here still has provisions for the emissions exhaust passage. This was the port flow tested, and the hole seemed not to affect the flow as the numbers delivered were strong.

The combustion chamber checked out at 58 cc, which with the deck clearance used, the head gasket volume and the volume contained in the D.S.S. pistons' valve cutouts, delivered a CR of 10.4:1 on our 331. About the time this Ford head was in the development stage, AFR's founder, the late Ken Sperling, was heavy into swirl and wet flow testing. This prompts the question as to how much of this head's success is owed to the possible incorporation of such technology.

The included flow chart shows the flow numbers we saw on our freshly calibrated bench. If you want to talk about numbers at 0.700 inch lift as is so often the case, then the intake produced 307 cfm and the exhaust 207. But that's not the real story. What seems to make these heads work is that both the intake and exhaust show excellent low and midrange flow. At low lift, this is very much a function of good valve seat design, and at midrange a combination of seat and port design. Even with less peak flow, a fat flow curve will usually outpower its lesser low-lift flow counterpart. Also, good low- and midrange flow has the ability to produce a power curve that hangs on significantly longer at the top end, thus advantageously delaying the shift point until higher rpm. The fact that the curves have largely leveled off above about 0.600 lift shows that the ports are well matched size-wise to applications where intended valve lift is in the 0.625 to 0.700 range. If a port continues to show increased flow much above the maximum valve likft used, it indicates the port is too big for the application.

Step By Step

View Photo Gallery
Mmfp_0803_17_z DSS_347_stroker_engine_build
The lash was set, and the rocker girdle was attached.
Mmfp_0803_18_z DSS_347_stroker_engine_build
Our D.S.S. 331 needs to breathe, so we chose an Edelbrock Victor Jr. manifold.
Mmfp_0803_19_z DSS_347_stroker_engine_build
With the manifold secure, we slid in the MSD distributor and topped the engine with a BG 750 carburetor.
Mmfp_0803_22_z DSS_347_stroker_engine_build
With the engine complete, it was on to the dyno.
Mmfp_0803_23_z DSS_347_stroker_engine_build
A variety of collector lengths was tested.

In addition to flow, we also measured the swirl. The fact that this proved to be very good (uncharacteristic for a conventional style Ford head) is a strong indicator as to why these heads have historically produced such a wide power curve. Based on what we found on our bench, it's hardly surprising these heads work as well as they do.

The Valvetrain
It should go without saying that we'll use roller rockers for this engine. Although the Comp Cams items selected are far from exotic, the combination used did take some care during assembly to ensure it would all work right. With the springs being used, the forces are such that we are moving into 3/8-inch-diameter pushrod territory. However, there are some space constraints that make this a doable but less-than-easy move. What saves the day here is that the short 5.0 block height also means a shorter-than-usual pushrod. This being the case, we can just about get away with 5/16-inch pushrods. To get the rockers to align over the valve tips, it's essential that the guideplates be positioned to allow this before having the rocker studs tightened down, as alignment only occurs over a small range.

Once all the rockers have been installed and aligned, it's time to fit the stud girdle. The obvious intent of the stud girdle is to add rigidity to the rocker system. We can safely suppose it does that, but the best part of having a stud girdle is that it allows the rocker adjusting nuts (polylocs) to be clamped in place rather than relying on the socket locking screw normally used for this purpose. By employing the stud girdle to clamp the adjusting nuts in place, a more accurate lash setting can be achieved. Lash for our valvetrain was set to 0.018 inch and 0.020 inch for the intake and exhaust, respectively.

Intake Manifold
Our first move here is to check the amount of space for the valley end seals to fill. With some head/intake combos, this can be a larger gap than the thickness of the regular cork seal. This was the situation for our heads and intake combo, so the cork gaskets from the Fel-Pro set were used together with a bead of sealer to make up the small difference.

Those of you who have explored a number of cylinder-head/intake combinations have more than likely come up against the problem of such a combination achieving a respectable port match. It's not uncommon to burn up four hours or more to match an intake to the heads because this is so often a necessary move. With our build, things were fortunately a little different. Our Edelbrock Victor Jr. was a near perfect match for the ports in the AFR heads. The only place a mismatch occurred was in the corners. Here the heads had a larger corner radius than the runners in the intake manifold, but the difference was small. This being the case, it was left as is.

Ignition And Carburetion
Ignition on our 331 was to be handled by an MSD system from distributor to spark box. To make wiring up the plugs a neater deal, it helps to install the distributor first, then the carb, followed by the headers.

The carb used here was a 750 Mighty Demon from Barry Grant. It's not often appreciated, but these Demon carbs' flow rating is with wet air, not dry, as is the case with most other carbs. The 750-cfm wet works out to about 810-820 cfm dry, so that's the size you need to relate to here. This may seem like a lot of carburetion capacity for an engine of only 331 inches, but there is justification. The higher-than-average booster gain on these carbs means they are not overly sensitive to dropping low-speed output if the carb is sized more for top end power. We also have some deep-breathing heads and a valvetrain with enough lift to access the heads' high lift flow. Our thoughts were that although BG might recommend a 650 for this application, we felt a 750 was worth trying. If it didn't work, the dyno would show it, and the fix would take only five minutes at most.

Step By Step

View Photo Gallery
Mmfp_0803_21_z DSS_347_stroker_engine_build Mmfp_0803_24_z DSS_347_stroker_engine_build Mmfp_0803_25_z DSS_347_stroker_engine_build
Accel wires bring fire to each hole, while Kooks headers expel the gasses.

After the carb went on, the headers were installed so as to allow a plug cable routing that best clears the ultimately hot pipes. From here, it's just a question of routing the plug cables in the neatest manner possible. As far as plug-cable routing goes, our project engines need to look good as well as perform. As of late, a completely different strategy has been used since ACCEL introduced its 10.8mm sheathed cables. Normally on our engine builds, a cable support system would be used for looks and to separate the cables so that big plug gap wouldn't cause voltage to cross-fire from one cable to another. The extra 1mm wall thickness of the 10.8 mm cable over the 8.8 offers more insulation than 1/4 inch of air gap. What this means is that the cables can now be bundled and tie-wraps used to hold them in place. This greatly simplifies wiring up a functional system, and it looks really good too. With the plug cables done, our 331 D.S.S.. track-day engine is looking good and ready to ride our DTS dyno.

Dyno Time
The nice thing about having an engine cart-style dyno like a DTS is that you can load the engine on the cart outside of the cell, wheel it in, and hook it up in a couple of minutes. With the timing set to 32 degrees (total advance), we gave the 331 a two-hour break-in on Castrol GTX.

After a post break-in service using Joe Gibbs racing oil (roller grade as used in their Busch cars), the first exploratory pulls were made. Two things were quickly learned. One was that the engine was quite fussy over the plugs it liked best, and that the carb calibration required for this small-block Ford was way different than the last engine on which it was run (393-inch street Windsor). What started off looking a little disappointing ended up as a "Wow-we may have too much power."

Getting to this point took not only carb jetting and ignition tuning, but also plug selection and secondary (collector) exhaust lengths. When we first hit the dyno, our 331 could only muster some 400 lb-ft of torque along with 420 horses. The jetting and ignition were no surprise, but the sensitivity to plugs was. We ended up with Autolite 3922s.

Our dyno headers are made by Kooks Headers of Bay Shore, New York. We use these for at least two reasons. First, they're extremely well made. Second, they're good for output. Our engine made its best overall curves when the secondary pipes were adjusted to 14 inches from the exit end of the four merged primary pipes. So how sensitive was it? Way more than we usually see. Just a 3-inch change showed as much as an 18hp difference.

As for the power, the two days spent on the dyno proved worthwhile. Because of the big cam, output much under 3,500 rpm was hardly sensational. As for idle, the lowest rpm with any kind of smoothness was about 1,050. When the dyno tach hit 4,000, you could hear this engine come alive. As you can see from the graph, we stopped testing at 7,300 rpm, and it looked as if that was still a couple of hundred rpm short of peak output. The reason for stopping short is that this D.S.S. unit was making more power than anticipated and, though it was selected, it was still nonetheless a stock-block.

An educated guess here is that, on the track, the best shift point would have been about 8,000 rpm. This poses the question of just how long the block would last. From the graph, it looks as if the power is still climbing to the extent that the 540 figure may have been surpassed. As for torque, a peak number of 442 lb-ft was seen, and that's a pretty respectable figure for a 10.4:1 331-inch engine. As for having too much power, we had an instant fix-a 7,000-rpm chip for the rev limiter.

Clearly this 331 was capable of making power to the extent that it deserved a significantly better block such as those made by Dart, Ford Racing Performance Parts, or World Products. If the budget had been about $1,200 more, that would've been the way to go. Not only would this have delivered the extra block strength needed, but also the fact that a bigger bore for 358 cubes could have been used. Even on pump gas, such a combination could well have yielded some 590 hp along with an 8,000-rpm redline.