July 23, 2006



Heads Under Pressure

Nick Harker

Christopher Tockey headsunderpressure@uidaho.edu

Peter Britanyak



| |

Dr. Karen DenBraven

Mechanical Engineering Professor

University of Idaho

Dear Dr. DenBraven,

Here is our interim design report showing our design work for the 2006/2007 snowmobile instrumented head. This report is for your review. Included in this report is all important models, design steps, and drawings created during the summer session 2006.

Our final design of the head utilizes a smooth asymmetrical combustion chamber to bring the sparkplug closer to the fuel cone and modified mating surfaces decreases overall height. It is also designed to work with new injectors and sparkplugs. The pressure sensors are located on the head behind the injector and can be removed and replaced with dummy sensors for use out of the engine testing bay.

We would like to thank you for giving us the opportunity to work on this project. From this project we have learned about client - team dynamics and the details incorporated in head and combustion chamber design.

Sincerely,

Heads Under Pressure

Instrumented Head for the Clean Snowmobile

[pic]

Team Heads Under Pressure

Submitted to:

Dr. Karen DenBraven

UI Mechanical Engineering Professor

Report submitted by:

Peter Britanyak

UI Mechanical Engineering Student

Nick Harker

Mechanical Engineering Student

Christopher Tockey

UI Mechanical Engineering Student

Table of Contents

1. Introduction

1. Background

2. Deliverables and Specs for Snowmobile Head Design

2. Modeling

1. Compression Ratio

2. Squish Velocity

3. Manufacturing Approach

4. Concepts Considered

1. Week 3 First and Second Iteration

2. Week 4 Third Iteration

3. Pressure Sensor

5. Concept Selection

6. System Architecture

7. Economic Analysis

8. References

Appendices

A. Squish Velocity

B. DFMEA

C. Gant Chart

D. Compression Ratio

E. Drawing Package

1.0 Introduction

1.1 Background

The stock Polaris Liberty 600cc is a two stroke naturally aspirated carbureted engine. With this as a platform, the University of Idaho Clean Snowmobile team sought to reduce emissions while maintaining drivability. In order to accomplish this, the first major change to the engine was to go to a direct injection (DI) setup. For the DI setup the cylinder head was redesigned to accept the fuel injector. The spark plug was also relocated so the electrode would touch the fuel cone. Direct injection allows the engine to run in two modes, stratified and homogeneous. Stratified combustion improves low RPM and light load fuel consumption and performance. During this mode fuel is injected late to preventing it from mixing prior to ignition. Homogeneous combustion is common in most engines and works well for high RPM and load. During homogeneous operation the fuel has time to mix with the air before ignition.

The first DI head design for the Polaris engine had a higher compression ratio than stock. This higher compression head experienced problems with detonation. The next design was a similar DI head that had a lower compression ratio to allow for the application of turbo system. The 2006 snowmobile ran the lower compression ratio head with natural aspiration.

To effectively tune the direct injection engine, pressure sensors in the head would provide cylinder pressure data. In-cylinder pressure can be used to identify cycle-to-cycle variations, peak pressures, identify detonation, calculate the mass burned fraction of fuel, and calculate the heat release of combustion. All of these parameters can be used to optimize combustion. The main goal of our design project is to design an instrumented head with pressure sensors to gather in-cylinder pressure.

Dr. Karen DenBraven, Andy Findlay, and Justin Johnson are our clients. Dr. Karen DenBraven is the advisor for the snowmobile team and our advisor for our project. Andy and Justin are graduate students that work on the snowmobile. Our project’s scope also includes many changes to the combustion chamber. Our clients laid down specifications for the new head design and our job is to meet them while retaining aesthetics and manufacturability. New sparkplugs and fuel injectors will be used to improve performance at high engine speeds. The difficulty has been to satisfy all the specifications.

2. Deliverables and Specs for Snowmobile Head Design

Deliverables

• Head Re-designed with pressure sensors for in-cylinder pressure data

• An analysis on whether casting the head is possible and cost effective as compared to machining

Specifications

• Pressure sensor to acquire in cylinder pressure data

o Design for Kistler 6052C sensor

• Sparkplug

o Design for NGK IZKR6B Plug

o Located 30º from the center of the exhaust port (current location)

• Injectors

o Design injector seating as per drawings (Appendix A)

o 5º injector spray angle towards the center of the intake port

• Combustion Chamber

o No dome offset

o 34% Squish area (same as turbo head design)

o Compression ratio of 6.5:1 (same as stock head)

o Dome radius .75 inches

o Spark plug electrode touching fuel cone

o Max squish velocity of 15-20 m/s at 7800 to 8200 RPM

o Squish diameter of 1.95 inches

o Squish radius .25 inches (double the turbo head)

[pic]

Figure 1.2

2. Modeling

In order to meet the design criteria TK solver was used to determine the ratio of squish volume to dome volume as well as overall compression ratio. Solid Works was used for solid modeling of the head and verifying the compression ratio and geometry. Master Cam will be used for programming the CNC mill for machining the head.

2.1 Compression Ratio

Compression ratio is typically calculated with the simple equation:

[pic]

The compression ratio was calculated with TK solver; the program is located in Appendix D. The ability to calculate the required clearance volume allows us to check the required displacement of the combustion chamber in our solid modeling. In addition to calculations we also filled the combustion chamber of the stock head with water and measured it to find the combustion chamber displacement of 36.3cc. This gives a stock compression ratio of 6.5:1

2.2 Squish Velocity

The velocity of the mixture as it is being compressed and forced from the edges of the cylinder to the center of the combustion chamber is called squish velocity. The squish velocity is a function of the speed and position of the piston, the relief and combustion chamber design.

Using the model outlined by G. P. Blair [1] the squish velocity was determined using values from our solid model. Calculations are shown in Appendix A.

3. Manufacturing Approach

We spoke to Berry Ramsay (President of D8) and David Deaton (Engineer at D8) about the possibility of casting the head. D8 is a local production facility in Potlatch Idaho, where they mostly create one-off molds for plastic injection. In speaking to them it seemed apparent that they would be capable of making a casting of the head, however there might be issues with porosity.

D8

Pros:

• After initial investment, cheap and quick to make.

• Easy “quick” prototyping

• Casting knowledge and experience gained to the clean snowmobile team

• Easier to implement design changes with cast head blanks

Cons:

• Expensive initial cost

• Out of house (less control)

• Time to create first head

• Unknown design criteria

• Still requires machining of final surfaces

Our research led us to other casting companies including Travis Pattern and Foundry, ECK Industries, and Lost & Foundry. Eck and Travis Foundry were not a feasible possibility due to the small number of parts that we would be casting and the associated cost and lead time. However, Lost & Foundry is a family owned business in Spokane, Washington with the goal of providing industry foundry techniques to individuals. They offer equipment at the cost of $400 that would allow for in house casting. This would allow for control over the process and timeline. Due to the lack of knowledge there would still be a significant difficulty to cast in house.

In house

Pros:

• Cheap

• Easy “quick” prototyping

• Casting knowledge and experience gained to team

• Easier to implement design changes with cast head blank

Cons:

• Would need to machine pattern

• Unknown design criteria

• Still requires machining of final surfaces

• Area for furnace and University approval

• Lack of knowledge and mentoring

Other concerns:

There are several physical differences between cast and billet aluminum. The heat transfer rate and the strength could require major redesign of the head for casting. Porosity in cast aluminum is also a concern. The pockets of air in the material could be machined into, leaving a cavity in a gasket seat or combustion chamber. Also an air pocket in a thin web or wall of the head would be a potential failure point.

Conclusions:

Industry uses casting almost exclusively compared to billet machined parts for production; because of this, the knowledge would be valuable to students for the transition to industry. D8 would be the only viable way to have a head cast; having said that and weighing the pros and cons, we have decided that machining is more feasible for manufacturing.

4.0 Concepts Considered

Initially the combustion chamber started out as a symmetrical revolved dome. The sparkplug to fuel cone distance caused problems with this symmetrical design. To have the sparkplug located on the surface of the dome and halfway between the injector and the squish makes the plug to fuel cone distance near impossible.

4.1 Week 3 First and Second Iteration

To bring the sparkplug electrode into the fuel cone the first attempt was to use a sparkplug with an unthreaded portion that could protrude into the combustion chamber. This was not sufficient to obtain the correct electrode to fuel cone distance. To get the sparkplug farther into the combustion chamber a boss was put in the wall to cover the threads of the sparkplug.

Figure 4.1

A symmetrical combustion chamber is preferred because it allows tooling to be made to cut the combustion chambers out in a single operation. The boss protruding into the chamber would mean that it would have to be machined with a ball end mill and would greatly increase machining time.

4.2 Week 4 Third Iteration

Andy and Justin recommended a flat side to the combustion chamber. It would have the same affect as the boss and would protrude into the combustion chamber far enough to cover the sparkplug threads. The advantage of the flat face instead of the boss is a very smooth transition face allowing for better flow of gases. The boss limits the amount of protrusion, where the flat plane makes it possible to move the sparkplug into the fuel cone while meeting specifications and not disrupting flows. Figure 4.2

4.3 Pressure Sensor

The pressure sensor also has different options for placement into the combustion chamber. The pressure sensor can be implanted into the sparkplug, in a mounting sleeve passing through the coolant passage or mounted by threads into a boss leading to the combustion chamber. The sparkplug mounted sensor was previously tried and there were issues with removing the sparkplug without damaging the sensor. Since we are redesigning the head a boss can easily be added for the pressure sensor so that it can be threaded directly into the head.

Figure 4.3

5.0 Concept Selection

Our final design uses an asymmetrical combustion chamber that improves on the D-shape by smoothing the transition in all directions which improves flow while meeting all specifications.

The outer mating surface will meet at the midpoint between the top and bottom planes. The boss for the injector seat and pressure sensor will pass through the cap and be flush with the top surface. A step in the boss will allow for an o-ring sealing surface at its base.

Figure 5.0

6.0 System Architecture

The use of Design Failure Mode and Effect Analysis (DFMEA) so far is not beneficial since we will not be able to build testable prototypes prior to the finished product. The process of creating a DFMEA chart did help realize potential problems that could occur. The DFMEA chart is in Appendix B.

7.0 Economic Analysis

|Head bolts from ARP M8x1.25 60mm |$55.74 |

|Machine Time (Approx. 100hrs) |$1000 |

|Gasket Material $1.00 per foot |$10.00 |

|Kistler Tooling |$372.40+S/H |

|Injector Tooling | |

Approximate time spent by the design team - 250 hours

8.0 References

Blair G.P. “Design and Simulation of Two-Stroke Engines.” Society of Automotive Engineers, Inc. Warren, PA, 1996.

Bradbary N.,Packard K., Spelman K. “Design of a Direct Injection Two-Stroke Engine For Low Emissions Operation.” 2003.

Bradbary N. “Retrofitting Direct-Injection and a Turbocharger to a Two Stroke Engine for Snowmobile Applications.” Masters Thesis Defense Paper, 2006.

Findlay A., O’Neil R. “Design of a Direct Injection Two-Stroke Engine

For Low Emissions Operation.” Design Report, 2004.

Appendix A

Squish Velocity

Appendix B

DFMEA

Appendix C

Gant Chart

Appendix D

Compression Ratio

Appendix E

Drawing Package

-----------------------

[pic]

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download