Back to work

I have put the project on the back burner for a while so that I could take care of personal business. I have been going through my notes and sketches to pick up where I left off and decided to not do that. Instead of concentrating on cylinder head design I am going to finish the cylinder block. This is my way of stopping me from getting ahead of myself. I will post my progress and future plans after the holidays.

Progress so far

It has been a while since my last update. I have decided to integrate a good portion of the actuator assembly into the cylinder heads to make the engine a little more compact. Most of my time has been spent researching combustion chamber design, direct injection technology, intake and exhaust port profiles, and how to organize them efficiently in the CAD model. I will post detailed information and screenshots in the next few days.

New approach

My design goal of making this engine block compatible with a Gen I Chevy small block engine has caused more problems than expected. The biggest problem is the hydraulic actuators for the intake and exhaust valves. The integrated position sensors are too long to fit in the optimal position. Every solution to this problem I have drawn up that will allow the engine block dimensions to stay within limits has a negative impact on the actuator design. I have began drawing up a new engine block without the design goal of being SBC compatible. This will increase the number of custom designed components necessary for the physical prototype. I feel that to properly design a working prototype I need to start from the inside and work my way out, not the other way around. Here are a few screenshots of the new engine block so far.

Position sensor problems

I was doing some trial fitting after downloading CAD data for the smallest MTS linear position sensor I found at this point in time. At first I thought it would be too big but I am going to spend some time trying to tweak it into position to see if that is the case before deciding to not use this sensor. If I can make it work out, the pushrod will be a part of the actuator itself. Increasing the bore for the position sensors also created clearance problems for the caps that will seal the actuator bore. This problem can be remedied easily but I can see that this entire area will be problematic to the water jacket plates regardless. I will try contacting MTS to see if they offer a smaller sensor. If they do not, hopefully I can get this one to work.

 

Position sensors and hydraulic solenoids

I found this position sensor made by MTS last year while working the original design. I really like them because of the accuracy, response time, and durability. This is a non contact sensor so wear over time is not an issue and it is easy to calibrate and recalibrate as necessary to compensate for thermal expansion as the engine is running. MTS will actually custom make sensors for companies, but I do not have the money to pay for this luxury at this point in time. The major problem with this sensor is its size. It will require a cavity about 2 inches in diameter. I would have to move the bore centerlines which means that the push rods would be drastically angled. I thought about using an offset on the actuators to line the push rods back up with the rockers, but I am not happy with the added mass to the actuators required. Getting this system to operate at at 6000 RPM minimum is already going to be a challenge so adding weight to the reciprocating assembly is not a solution I will accept. I checked out the MTS product line again and found this position sensor. It has a diameter of 1.1 inches which I can squeeze into place by increasing the bore and adding a little material for support.

I am having doubts again about where to place the hydraulic solenoids. I am thinking about mounting them in the valley again. Having remote mounted solenoids would be ideal but that creates more problems than it solves. Having 16 separate hydraulic lines running into the engine block adds too many potential leak points, adds unnecessary clutter in the engine compartment, and complicates the way fluid is routed internally. These lines would also interfere with the accessories at the front of the engine and the bell housing at the rear.

My next major obstacle is going to be a crankshaft position sensor. Accuracy to the half degree would be perfect but, I am going to need something that is at least accurate to the degree. I have quite a bit of work to do before I really have to worry about this so I will just see what's out there for the time being.

Getting close

The block design is getting close to done. The only major piece of the puzzle missing is the main web thicknesses. I have the bolt hole locations but without the thicknesses, there is absolutely no way a Chevy crankshaft would properly fit. I found that SEMA members have access to GM CAD data. It will take me quite a while to save up the money to pay for a membership and get a hold of this information. I have searched for hours trying to find this information on the internet with no luck as well. For the time being, I will ignore the fact that the main webs are not correct. I also need to add bolt holes, oil passages, skirt the block, add the water jacket plates, and add water/steam holes to the deck. I can then add and remove material where necessary for strength and weight reduction and begin FEA and CFD. The next major piece of the puzzle will the the hydraulic manifold and solenoids. This will be fairly simple to complete. The hard part is finding the right solenoid. I have not received a positive response from any company I have contacted yet. Next on the agenda will be high resolution position sensors.

Fluid routing

Routing the hydraulic fluid for the valve actuators through the block is taking a lot longer than I expected. I wanted to keep the solenoids out of the valley and away from the heat and vibration. This would allow me to service/upgrade the solenoids without tearing into the engine as well. The downside is that instead of one pressure and return line going into the block, I have to have a line for each intake/exhaust valve going into the block. I spent a few days trying to come up with an efficient way of doing this and came up with a bolt on piece that looked similar to the valve body in an automatic transmission. I decided to ditch this because it bolted on. Expansion, contraction, and flexing would affect the mating surfaces and numerous seals and gaskets would be required. This introduces to many opportunities for leaks in the system. Any failure in this system would be catastrophic so I want to eliminate as many possible ways for it to fail as I can. This will also be cheaper since I will not need a separate piece machined, o-rings, gaskets, bolts and washers. Judging from the many designs on paper, I will most likely have to play around with the actuator locations in order to have these hydraulic fluid passages machinable. Looks like stock pushrods might not be possible. Here are a few screenshots of the block. I have not done much since the last post. I have spent most of my time drawing different possible solutions on paper.

Coming into focus

I have been tweaking the block design to get it as efficient as possible in CAD format. I am working on a five year old laptop so I need to keep things as simple as possible in order to FEA and CFD analysis on it in a reasonable amount of time once I get to that point. While tweaking the design and dimensions I decided to factor in the water jacket plates last because I found myself working around them when there is absolutely no need for me to do that at this point. I got a hold of the dimensions for the main bolt locations which will give me an idea of where to place the main webs, but I am still going to hold off on adding them until I have all the dimensions I need. Here is what it looks like so far:

It doesn't look much different from the last update but there have been many changes made. I will be working on the actuators themselves today. Once a rough draft done, I will place them in the desired positions in the valley area and begin to add the necessary support structure for them to the block itself. These will be similar in design to linear hydraulic actuators. The biggest challenge will be for me to place them in this area along with the hydraulic fluid pressure and return bores, have them communicate with the main oil feed (in order to lubricate the cylinder heads via the pushrods), and making sure the hydraulic fluid and engine oil remain separated. I will be using the spreadsheets I have been working on to get an estimate on the necessary fluid passage diameters. Hopefully everything will fit with no problems and allow me to still use stock length Chevy pushrods.

Finally making progress

I didn't realize how hard it would be to get Chevy small block measurements. I am doing the best with the few that I have managed to get. I changed my mind again about going the cast or billet route. I have decided to go billet for a few reasons. The most important of these is increased strength. I would hate to have to pay to have a second block cast because of a crack or because of a defect introduced during the casting process. After doing a little research, I found that I can use water jacket plates to address the water passage problem with machined blocks instead of going to an open deck design.

Starting over from a "blank sheet" is proving to be a great decision. Ditching the parametric approach has simplified things drastically. As seen above, I have a "rough draft" of the cylinder banks completed so far. You can get an idea of how the water plates will attach to the block. The cylinder head bolt bosses extrude all the way down which means the cylinder head bolt holes will be blind. I decided to go with the Gen I head design simply because they are very cheap to obtain. Keeping the first physical prototype cost as low as possible is one of my design goals. I want to place the fluid passages for the intake/exhaust valve actuators where the camshaft would normally be located, position the actuators where the lifters are normally located, and hopefully be able to use a stock pushrod length. This would knock out the cost of having to order custom length pushrods. I have a few designs for the actuators on paper but have not decided which one I will try to incorporate into the block just yet. I am still trying to get the dimensions for the main web thicknesses and spacings before I add them to the CAD model. These must be accurate in order for this block to accept a Chevy crankshaft. I am going to concentrate on the valley area until I am able to either measure a block or get these measurements somewhere online.

Still at it...

I received a hand written letter from Hydraulics & Pneumatic Sales, Inc. today. It is nice to see a company that is taking the time to help me out. I have been working on the spreadsheets and have made a ton of progress. I am holding back on posting any of the numbers I have so far until I have everything factored in.


I have started the block design almost from scratch. I tried a parametric approach originally. This way I could adjust any parameters and the block would automatically update as required. This did not work well with the idea of using off the shelf Chevy parts. I also spent a lot of time creating equations instead of simply typing in specific measurements. I have decided to ditch all parametric equations and dimension the block to accept a Chevy Gen I style crankshaft, rods, and pistons. I am going with a 4.001" bore and 3.48" stroke for a traditional 350 cubic inches. Horsepower and torque goals are 626 and 548 respectively at 6000 RPM, good enough for a mid 10 second Camaro. I will not know if 6000 RPM is a realistic speed goal until I finish my spreadsheet, but from what I have so far it will definitely be achievable with the solenoids I am looking at. The torque curve will be a flat line if this system works as intended (at WOT since volumetric efficiency is intentionally cut back for throttling). Since horsepower is a function of speed and torque, if I can achieve a higher engine speed then the power should increase as such (ex. 782 hp and 548 ft/lb at 7500 RPM, knocking on 9 second's door in a Camaro). This is dependent on the valve actuator's ability to maintain 100% or better volumetric efficiency. These numbers are if the engine has an 80% mechanical efficiency, 100% combustion efficiency, 46% brake thermal efficiency, and a BSFC of 0.30. This is more than likely not going to happen, and I still need to finish the spreadsheet and factor in a lot more variables. I said I wouldn't post any numbers but I did it anyway...

I am going to get the lower block dimensions the best I can without being able to take any physical measurements tonight, talk with the hydraulic supply guys tomorrow, and decide if I should design the deck to accept a Gen I or Gen III/IV style cylinder head. I have decided to ditch the goal of designing this to be machinable as well. This will allow me more freedom in design and will possible be cheaper to have this block cast instead of machined, as well as being able to ditch the open deck for added strength and reliability. These decisions will bring the overall cost and development time down.

First Response

Parker responded to my request today. Turns out the solenoid I was interested in has been discontinued. I am waiting to hear from Clean Air Power. As of today, the solenoid that they make has been the best choice for this application. I have been drawing various  ways to integrate the hydraulic manifold into the engine block and am leaning towards a different approach. i wanted to integrate the entire system into the valley of the block. I am now thinking that it would be better to only place the actuators in the valley, have the hydraulic manifold located remotely, and connect the two with a "universal" hose. This way the solenoids are away from the heat, can be easily accessed, and the entire hydraulic manifold can be serviced, or replaced with an upgraded design, without having to disassemble the engine itself.

A friend of mine told me to look into 3D printing. I did and realized that it would be a great tool to have access to in the near future. A 3D printer is basically a machine that can sculpt 3D models out of ABS plastic (or other materials) layer by layer. In other words, I could design something, take the CAD model and "print" it, come back in a few hours and the have the physical object. I could make a block out of ABS plastic to test for fit and function or any design errors before even going to the machine shop. I could even make my own parts (intake manifolds, runners, electrical connectors, brackets, covers, housings, etc.) in house as soon as I need them if they can be used as ABS plastic. More expensive 3D printers can make circuit boards, multi part assemblies, and even color objects as it makes it. There is a do-it-yourself open source kit available that can be put together for around $500. Mendel RepRap

Still Waiting

I haven't gotten a reply from Parker yet. I requested product information from a few more suppliers. While searching for suppliers who could possibly help me out, I came across this solenoid:  Solenoid Valve. This solenoid would open in 72 degrees of crankshaft rotation at 6000 RPM and close in 108. I am still working on the spreadsheet and have to factor in pressure and flow along with these response times to get an estimate on total intake/exhaust valve travel, but needless to say, I will most likely get the best performance using this solenoid valve. This solenoid does not come in a 3-way/3-position configuration that I would prefer. I would be forced to use two solenoids per intake/exhaust valve instead one only one to compensate for this. This will add to cost and weight, as well as make the system take up twice the space.

I need to know what solenoid valves I am using to design a hydraulic manifold for it. I will not know the physical dimensions or port locations until this is done, so I can not design the mounting provisions onto the block itself yet. While I create my spreadsheet and wait on replies for the suppliers I have contacted, I will work on the main girdle design, main webbing, main bore surfaces, and decide where to locate the main oil supply. Since I plan on using Chevy crankshafts I need to find some measurements online to make sure my web thickness and spacing is correct. It would be nice to have a crankshaft to measure but I can not afford to buy one right now.

Hydraulic Solenoid Valves

After deciding to find a solenoid valve to use so that I could get its dimensions in order to determine how all of them would fit into the valley of the block, I began doing a some searching. In order to limit the number of solenoid valves needed, I would have to use a 3-way/3-position design such as this one: EMBV-09-3J.

Unfortunately I can not use this solenoid because of its 50 ms (0.005 s) response time. At 6000 RPM the crankshaft is turning 360 degrees in 0.01 s or 1 degree every 0.0000278 s. The crankshaft would theoretically turn 5 times before this solenoid valve would allow an intake or exhaust valve to fully open and close even once.

I have found that Parker has design that is much faster: DFplus. I have tried searching Parker's official website for CAD data and performance specs but can not find any. I have sent them an e-mail requesting this information and hopefully I get a response soon. In this link above it is stated that it has a response time of less than 3 ms and I have found a brochure stating that this time is 10 ms. I have even searched for response times of solenoids and found that one is stated to open in 5 ms and close in 2 ms. Unfortunately, even though those numbers showed up in the search results, the link was dead so I could not get the make and model of it.

At 6000 RPM, 5 ms equates to 180 degrees of crankshaft rotation and 2 ms is 72 degrees. This means, theoretically, that the solenoid valve would be off its seat for 252 degrees (roughly 7.0056 ms) if it was closed immediately after fully opening. I would have to use more than one solenoid to control opening and closing the intake or exhaust valve to make 6000 RPM a possibility if these were its response times because a 3-way/3-position solenoid valve would have to actuate twice per intake or exhaust valve action. A 3-way/2-position and a 2-way/2-position solenoid working together would be more ideal with these response times.

I am going to put together a spreadsheet to make determining the relationship between a solenoid valve's response times vs RPM much easier. That will give me something to do until I get the information I need.

Getting Started

I wanted to post some screenshots of the progress I have made but I can not get them to come out right. As soon as I figure out how fix this problem I will post them here.

I have started on the block. I have chosen to go with a bore of 4.000", bore spacing of 4.400", and a deck height of 9.500" so that I could use of the shelf Chevy pistons, connecting rods, crankshaft, main and rod bearings, rings, wrist pins, bushings, and locks. When the time comes to have the first initial block machined, it will be much cheaper to use the readily available parts to test the block instead of having to custom make them. I can also focus on the solely on the block and not have to design and test too many parts at one time. Once I get it to where I like it, I can then start working on a crankshaft, connecting rod, and piston.

The block at the moment is just a simple "rough draft". It is just two banks of four cylinders connected to the main webs. The banks themselves are just the dry cylinder liners, surrounded by an opening for coolant flow and then the outer walls. It is an open deck design. I chose this because I wanted the block to be able to be machined from a chunk of metal to avoid casting and the flaws that come with it. I am thinking of adding an insert of some kind at the top to address the known sealing problems associated with open decks.

 The block is skirted and I am designing it to use a one piece front and rear seal which will be a part of their respective covers. I plan on going with a one piece girdle to secure the crankshaft instead of caps. I have been researching solenoids so determine which ones I will most likely use. This way I can download the CAD data for that particular solenoid and begin to design the valve actuator system into the "V" of the block. I am choosing to place it here instead of directly on the cylinder head for a few reasons. The most important of these being the use of push rods and rocker arms. I can tinker with rocker ratios, line pressures, and orifice diameters to compensate for the solenoids "slow" open-close times in order to raise the RPM limit. Lower center of gravity and a more compact engine are also benefits of placing the system in this location. I would love to avoid using push rods, but at this moment the pros outweigh the cons.

I have been thinking of designing a manifold that the solenoids attach to and attaching this to the block itself. This way I decide to go with a different solenoid anytime in the future, I can machine a new manifold instead of having to machine a new block to accept the new solenoids. One concern of mine is the heat in this area and accessibility. It would be a pain to have to partially disassemble the engine just to check a solenoid.

Here is a .pdf of how it looks so far. Just a rough draft of the block, girdle, front cover, and cylinder liners. Once I figure out how to fix the screenshot issue I will post pictures directly here.

Bottom End-1.pdf

Introduction

A few years ago while attending Nascar Technical Institute, I came up with an idea for a bolt on that would act as a camshaft replacement for the small block Chevy V8 engine. It was simply a group of hydraulic valve actuators in a housing that would replace the existing valve cover. They were operated by solenoids that were controlled by a computer. I drew the basic design on paper and about a year later I was ready to build it and see if I could make it work. Once I received price quotes from machine shops and manufacturers of components, I realized that I did not have the money to build my first prototype and development came to a halt.

I was talking to a friend not too long ago and he gave me the idea of trying to get investors to fund the development of my idea. I began doing research to see what I needed in order to create a professional presentation for potential investors. It was not long until I came across the subject of digital prototyping. I was amazed that software existed that would allow me to design, build, and test my idea on a computer without spending a penny. I immediately began designing a digital prototype and testing it.

Months went by as I continually tested the design, researched parts and materials I would need, and updated the digital prototype accordingly. After I was happy with the fit and function, I wanted to place my virtual invention on a virtual engine to validate the design better. It wasn't long until I went from trying to create a digital small block Chevy for testing purposes, to designing by own engine block. Back to the pen and paper...

I decided that since I can design whatever I want and build it digitally, why not make a complete engine that is designed from the ground up to use my valve actuator idea. While i am at it, why don't I incorporate a lot of other ideas I have into the design. Intake/exhaust valves actuated by a computer instead of a camshaft, an oil pump capable of building up oil pressure before cranking, combustion chamber pressure senors, a Linux base ECU capable of monitoring engine performance and driver inputs, then adjusting valve timing, spark timing, fuel trim, etc. all while being fully accessible to users for modifications, monitoring, and reprogramming.

I have many drawing of how various components of this engine will work. I believe I have enough to began creating the digital prototype of the first version of this engine. I tried searching the internet for similar projects but I was surprised to only find two. Both of those were no longer active. I decided to create this blog to document my progress, share my ideas with people who have similar interest or projects, and to hopefully get some help designing this thing. It is my dream to day be able to have this engine installed into a Camaro so I can cruise the streets with a big smile knowing that I designed and built it!