School us on engines...

Discussion in 'SN95 4.6L Mustang Tech' started by GDawg, Dec 23, 2006.

  1. How much boost is needed? Intercooled?

    How much N2O HP? For low levels no changes. For medium levels you can go with some head work or just get a cam with the correct LSA and increased exhuast specs which is what most people do. Large amount you need a person who builds N20 engine for a living.
    Edit: Remember unless is a strip only car you are dealing with to different engine needs. With and without N2O. What will help the N2O can reduce the NA HP.

    I have never seen flow numbers where the entire motor was used. Normally the intakes or flowed using a radius entry but I have seen where the intake manifold was used for flow testing. On the exhaust side there or those that flow with no pipe and those that use a pipe. In most cases using the pipe will cause the exhaust numbers to be higher. If the person used a pipe you want to look at the diameter and length he used as this will change flow numbers as well. When flowed the head is bolted to an adapter which should have a diameter near the bore of the engine you are building. Again the bore size can change flow. Also check at what depression the heads were flowed normally in inches of water. Today most are flowed at either 25 or 28 but some use higher and older number were normally flowed at less depression. If you know what depression they were flowed at it can be converted to another depression.
  2. When it comes to FI (forced induction), it still mostly boils down to the pressure differential at the throat or more importantly the back of the valve. When you generate boost and thus create a relatively high pressure region, you have a greater pressure differential between the back of the valve and the inside of the combustion chamber. There are basically three ways that FI increases power.

    One, during valve overlap the scavenging process still works the same as N/A, but with greater force. Let’s assume we are at TDC with both valves closed and the spark plug just fired a brief moment ago. Now as the piston approaches BDC the exhaust valve begins to open to relieve the high pressure within the cylinder as a result of combustion and the end gas that is still continuing to burn. Gasoline does not explode like some think, it burns slowly. Compression just exacerbates this controlled burn to such a magnitude that it is mistakenly perceived as an explosion. Anyway, as the piston starts to come back up on the exhaust stroke, the intake valve is opened about 50°-75° BTDC on a street/strip motor depending on the cam profile. As soon as the intake valve is lifted the exhaust valve is still partially open and the hot expanding and still burning exhaust gases are rushing out of the exhaust valve very quickly. This provides a relative low pressure zone (pressure differential) at the top of the cylinder and actually helps to draw more of an a/f mixture into the cylinder by way of the intake valve. This is valve overlap. The more powerful the combustion process, the greater the ability of the scavenging process of overlap. Kinda like a vicious cycle but good. When you “stuff” more air into the cylinder with boost, you obviously have a greater combustion process and therefore a greater ability for scavenging during overlap.

    Two, when you generate boost and apply greater pressure to the back of the valve, you are compressing the air and the suspended fuel that is waiting behind the intake valve. This creates a relatively high pressure region behind the intake valve. Continuing on from above, once the intake valve opens, this increased pressure differential allows the compressed air/fuel to flow into the cylinder with much more force and therefore “stuffs” the cylinder with more air/fuel.

    Three, like mentioned above boost compresses the air and suspended fuel behind the intake valve. Whenever you compress a gas and a vapor suspension within a given and fixed volume (port volume) you increase the available mass of oxygen and fuel within the port (venturi). So when the intake valve opens, not only is it being rushed in with a greater force, but it is also a much denser mixture and therefore gets more oxygen and fuel into the cylinder.

    To optimize horsepower in an internal combustion motor, you want to have your overall ignition timing set to such that the resulting "normal" combustion process makes peak cylinder pressure at around 14°-15° ATDC, 16°-17° for an all out drag race motor. Obviously, you don't want peak cylinder pressure to occur before TDC. That excess heat and pressure would apply abnormally excess pressure on the piston during the compression stroke and place undue stress on the wrist pin, rod, top ring land, and can severely scuff the skirt. All not good.

    When cylinder heads are prepped specifically for FI applications, there are some differences from that of a N/A head. For instance more attention is placed on making sure that you do not have excessive tumble or vortices. On FI heads you have to also work them so that you all but eliminate boundary layer shear. You hear lots of people toss around the term “Port and Polish”. You never want to polish the intake runner, but back in the day on carbureted motors, engine builders would polish the exhaust port on the head. This was simply done to help eliminate carbon buildup inside the exhaust port. On EFI motors, polishing the exhaust port is a complete waste of time and money and is actually rarely done if at all. It has turned into more of a marketing gimmick much like blueprinting. The reason you don’t want to polish the intake runner is because you would disturb the boundary layer and create shear. This would disturb the laminar flow and would result in increased Reynold’s numbers/ratios. The little ridges that are left inside the intake runner after CNC porting are purposely machined. It keeps the fuel in suspension and promotes flow/velocity within the port.

    That’s about the best I can explain boost and timing without going off the deep end.

    Stan pretty much summarized flow testing. Normally it is done with an acrylic cylinder fixture at a depression of 28.0” of H20. The length and bore of the bore fixture will affect the observed flow. Just the same, the dimensions of the exhaust pipe used will affect the exhaust flow. There are lots of different types of flow benches out there, but the SuperFlow products are widely recognized to be very accurate and professional. For example you have the SF-600 which can test up to 600 CFM or 240 HP, more than enough for most street/strip motors. Then you have the SF-1020 which can test up to 1200 CFM or 240 HP for the more radical race engines.

    Sorry for the book. :)
  3. Here are some links to a pictures on Larry Meaux site.
    An intake manifold being tested with the head.
    Here is a number of pictures of different heads and combinations on a flow bench.

    Also the flow bench is a base and can be added to. Other tools can be used with it like swirl meters tumble meters and computer data capture systems. Check out Audie Technology
    The hot new thing is wet flow where a colored liquid is used. Either digital pictures or video is taken for study. Kevin do you have any wet flow pictures you would like to share?
  4. I wish we did, :D. The whole corner of the shop was green, including me. We had no idea what we were doing at the time, so I had just talked to a few friends of mine in Las Vegas that had done this before with their own setup and they gave me a few pointers. At the time, we did not know about the white Dykem that was available to contrast the green or red dyes used to look at under the black light, so we were allowing way too much dye to be pulled through and well, we made a big mess.

    Now people like Superflow and Mondello make some really nice wet attachments and recovery systems. There are probably some pictures on the Superflow website. Maybe one day when the demand creates the need for one.
  5. Thanks for the responses.

    So if I wanted to add a supercharger to my car, than which aftermarket head would I want or would I just stay stock. I want a street/strip car, something making under 500rwhp. I have been looking at getting new heads such as MPH stage 2.5 heads or something like that. Something pretty much off the shelf. Are there heads or similar heads good for FI. Would I even need PP and heads for FI. I understand about the intake vavle not being polished, too much off a good thing either way can be bad (too much turbulance or not enough), but you mentioned vortices and boundry layer shear. How are they detected? Can you acutally measure that or detect that with a flow bench or is that calculated from the data from a flow bench.

    One more thing. My first car was a 1993 Mustang LX 4-cylinder and it had 2 plugs/cylinder. I know the theory behind why to use 2 plugs/cylinder, but is that not a normal thing for the modern engine and what is the reason for not doing this? Does it not add enough power/cost or is not enough room to add another plug? Do engines still use this type of setup?

    You can get into as much detail as you want or give me suggestions to books/articles/websites with information. I won't become an engine builder, but I am an engineer and I love to learn about this stuff, plus I love cars. So, yes I am an enigineer nerd. I read engineering books, and I love watching the discovery channel.
  6. It's not the intake valve that you don't want polished, it's the intake runner (port) itself that you do not want polished. Most of the CNC'ed heads out there are not necessarily specifically ported for nitrous, FI, or N/A. The CNC programs are just developed off of a master cylinder head that has been hand ported or ported by another CNC machine. Then that master cylinder head is copied by a 5 axis CNC program. The CNC machine will then replicate the port dimensions that it has been programmed to copy. So in many cases the CNC'ed head is only as good as the master port that it is copied from. You will have to specifically ask for a cylinder head to be setup specifically for FI and it will usually cost a little more money as it will normally have titanium valvetrain components and better springs and will take a little more man hours to get the ports just right.

    Yes, tumble motion, swirl motion, shear, radius velocity, boundary velocity, etc. can all be measured to a certain degree on a flow bench. You just need the proper tools and/or equipment. There are swirl meter and tumble fixtures available to use in conjunction with a typical flow bench. Usually on 2 valve heads, you will want to focus on a positive swirl motion. On 4 valve heads, you will want to focus on a positive tumble direction. In the past I have always played with both to get the most out of a head.

    I believe you are probably right about the 2 plug per cylinder setups. On motors with larger combustion chambers, high HP motors, and especially hemispherical combustion chambers the 2 plug per cylinder actually provides a great benefit. But, on your average street car, I don't think the performance nor fuel economy/emissions benefit outweighed the doubled cost of adding twice as many plugs.

    I asked a friend of mine at HMS if he had a picture of a swirl meter or tumble fixture. He sent me this picture of a swirl meter:

    View attachment 408711
  7. It is very important to point out CNC'ed head is only as good as the master port that it is copied from . Some people see CNC and think it magic. The only real advantage of CNC porting over hand porting is cost. The grinding of a head for the customer is just the end product of someone spending many hours on the flow bench with a probe deciding where to grind material over or maybe adding material like epoxy in area. Note epoxy can only be used in intakes exhaust have to be welded.
  8. I couldn’t agree more. It still never ceases to amaze me that lots of people still think that CNC’ed heads are the end all be all to cylinder heads. When you clamp a head down into the CNC machine, you are totally relying on Ford's tolerances as to where they cast the ports, valve seats and guides in relation to the outer dimensions of the cylinder head. If Ford’s tolerance on these 2 valve PI heads is just .010”-.012” in either direction, that means you have to program the CNC machine to leave up to .024” extra material in total so that you don’t end up getting the port walls too thin or worse, hitting water. The CNC has no idea if its .001” too close or .001” too far away from the water jacket or that valve guide boss. The cutter only cares about going where the program tells it to go. There are even spots within the intake runner that the cutter cannot reach. Then you have the human variances in the clamping process and the casting variances. So all this considered, you need to always setup the CNC program conservatively so that it doesn’t ruin a head.

    I have never come across a CNC’ed cylinder head in the past that could not be improved upon by hand. CNC’ing is a great piece of technology for producing large runs of cylinder heads to within as little variance as possible, but the CNC process is not the final say so in heads that will perform.
  9. A picture from a wet flow bench.
  10. Do you guys think that Ford will ever do away with camshafts and use a solenoid actuated valve like some exotics and the new C class Mercedes are already using? What about a setup like the F1 motors use called the PVRS(pneumatic valve return system). What about rotary valves or HVA(Hydraulic Valve Actuation) systems?

    The 42-48 volt systems that would power these systems is also said to provide a significant reduction in weight from not having to run as thick a wire for all the harnesses in the car.
  11. ^So they rely on an electric motor basically to open the valve? Im not familure with that at all....but it does make sence. Im also sure that it could be retrofitted fairly easy. Would be nice to essentially put any "cam" into your motor by just jumping on the lab top and altering the program :D . That would be sick!
  12. Yep, just imagine having complete control over your valve timing. They will probably come out with some sort of proprietary programming that can dynamically control the actuators. I think they should make them a solid state type actuator so that durability wouldn’t be an issue. Also having electro-pneumatically actuated valves would all eliminate the need for a throttle body. They might still use a TB for the first few model years just as a precautionary measure. Our idles would be smoother, fuel economy would significantly increase, and best of all we would make a lot more power not only from the freed up power of not having to turn the camshaft and compress the springs, but the ability to produce a nearly square wave lift profile throughout the nominal RPM range. Another thing that PVRS can improve upon is valve bounce. Spintron results have shown that the valve actually bounces off the seat many times before it actually comes to a rest on the seat. This kills power. PVRS can be programmed so that the valve is set down “softly.” It has been said by a few of these companies that they will utilize a special software algorithm to control the actuator coil currents so that the valves are decelerated to a speed near zero as they land, in conjunction with a switching time of barely three milliseconds

    An electronic valve timing (EVT) approach is described which utilizes high-speed, electro-pneumatic actuators producing nearly square wave valve lift profiles throughout the nominal engine speed range. The resulting impact on engine "breathing" combined with the controllability afforded by totally programmable valve timing produce an engine with performance capability radically different from any traditional internal combustion engine.

    Here is another example of a slightly different approach.

    Even though this Coates approach does not currently allow for the degree of control possible with computer controlled direct actuation, it’s a big step forward as compared to the current poppet valve technology. A 5.0L stock engine from a Lincoln which produced 260/249 went to 475/454 when equipped with their head and valve system.
  13. I believe as cost per unit drops then we will start to see it. Remember the time we are talking here for these thing to operate in.
    6000 RPM 1 crank rev 10.0 millisecond
    9000 RPM 1 crank rev 6.6667 millisecond
    and that F1 engine at 18000 RPM 1 crank rev 3.3333 millisecond
    With the above times and lift to produce a nearly square wave lift profile throughout the nominal RPM range I do not want to see the acceleration rates.

    On the Coates site
    "The Coates Spherical Rotary Valve Engine is the most advanced in the world, with the most positive valving system ever built. The breathing capabilities of the system are almost double that of a poppet valve. For instance: a static test of a five-litre poppet valve engine on an airflow machine produced a reading of 133 cubic feet per minute (CFM) with valve fully opened. The five-litre Coates Spherical Rotary Valve Engine on the same machine, however, produced a reading of 319 CFMs fully opened; a colossal advantage in airflow comparison. A five-litre poppet vavle engine tested on a dynomometer under the same loads and conditions at 5500 produced 480 BHP and 454 foot pounds of torque. The maximum RPMs on the poppet valve engine were 5700 RPMs; the Spherical Rotary Valve Engine in comparison reached 14,850 RPM's,"

    Are they saying in the same head they replaced a poppet valve with a rotary valve
    and when from 133 cfm to 319 cfm on the flow bench?
  14. Supposedly Siemens Automotive is claiming to have their switching time down to 3 milliseconds. But they are going to have to use sensors with an extremely high resolution in order to properly monitor the coils/actuators and give feedback to the system. The acceleration forces involved are going to indeed be lofty. With our conventional cam/valve spring setups, the acceleration/force/valve lift curves are nearly linear. As we get closer and closer to a perfect square lift profile, the acceleration/lift curves will become much more complex.

    Theoretically these digital switching circuits can achieve a true square, but I think the best these solid state actuators could do in reality is a triangular wave because of the harmonics that are inherent to the engine running. Maybe we could fill the actuators somehow with a glycerin/silicon solution to help eliminate as much of the harmonic influence as possible.

    I think Coates used their own proprietary cylinder head design that is used in conjunction with their valve system.
  15. Being in the computer field since the late '60s and watching hard disk actuators both hydraulic and voice coil get faster and faster and moving large heads, this was only a matter of time and dollars.

    Thanks that makes more sense.
  16. I would love to see one of these setups on a wet flow bench. I would like to see what everything looks like in the combustion chamber when these valves instantly open and instantly close. I would also imagine they are going to have to use Titanium valves or something else that is very light.