# Rotax 447 muffler how it works

### Help Support HomeBuiltAirplanes.com:

#### cptcliffhanger

##### Well-Known Member
I ran my 447 over the weekend and was impressed to see how quiet it is.. I wonder if anybody knows/understands the workings, and perhaps recipe for the rotax mufflers we see on nearly all 2 stroke rotaxes.. I'd like to be able to use the recipe on different 2 stroke engines if anybody can describe how displacement and RPM etc.. affect the internal sizing of chambers and tubes to recreate such good power band and silencing as this:

#### Armilite

##### Well-Known Member
Rotax's use a Header Megaphone Type Exhaust for a Small Gain 1-3hp, goes from Small to Large(Megaphone) which then the Gases run into the Muffler! All Mufflers work the same, to Slow down the Hot Exhaust Gases to let them Cool. The Gases come in Chamber 1, hit a dead-end wall, bleed off into Chamber 2, then Bleed Off into Chamber 3, then Out the Exhaust where it hits the much Cooler Air. That Hot Gas is what makes most of the Noise over the Normal Engine Noise. Sound is rated in DB. Download one of the Free Sound Meters Aps for your Smartphone. A Suppressor for a Gun Works the same way.

See Through Suppressor in Super Slow Motion (110,000 fps)

A Tuned Pipe uses Sound Waves to Push Back on Raw Fuel/Air going right out of the Exhaust that can Gain you up to 45% more hp!

Sound Meter

2 Stroke Exhaust Calculator for making a Tuned Pipe. 2 Stroke exhaust calculator - Apps on Google Play

#### koyama

##### Active Member
The can that many people call the muffler is actually not a "muffler". It is actually a broadband expansion chamber. If you were to remove it, the 447 is nearly impossible to get started, and it will produce about 20 HP give or take. Most people that are used to 2 stroke motorcycles do not recognize them as an expansion chamber for 2 reasons:
1) They are 50% the size of a single cylinder chamber, as they are designed to provide a reflection to the OPPOSITE cylinder via the Y pipe.
2) They are not designed to 'peak' the torque curve as are most land based vehicle 2 stroke pipes.

Some details:
Two stroke exhaust pipes generally have specific sections:
2) Diffuser (divergent zone)
3) Belly
4) Baffle (convergent zone)
5) Stinger

Note: Some shops and engineers use different terms than the above and even have more sections defined. In the early 80's I wrote a software package for designing expansion chambers. I used different terms and more sections because of the way that the software did its calculations and derived reference points, (no other reason). As it turns out, this was a bad decision as it ended up causing many people to rename the parts, and I still hear the terms I defined in my software used in the real world at times... That was NOT my intention...

On the 447 and other similar Rotax engines, the header is the Y pipe and the elbow immediately after (if equipped). The diffuser is the tapered pipe between the coupling and the can. The belly is the center section of the can. The baffle is the tapered tube inside of the can AND the rear section of the can. The stinger is the pipe that exits the can. It is basically a folded design, so it is harder to recognize it as an actual expansion chamber.
Here is where things get really interesting. The big holes in the front end of the can in the sides of the diffuser are where a good portion of the actual tuning happens. These holes, both the size and quantity are critical, and it is in many models the only thing that changes between engine models. (IE the 447 and the 503). What happens here is that the faster the engine turns, the smaller the chamber appears to be to the traveling pulses. You can think of it as a cup with holes in the bottom and sides. If you pour water in slowly, it is mostly empty (large volume of available space inside). If you pour water in faster the cup will fill up more (less available space inside). What this effectively does is make the chamber appear smaller as the engine turns faster. Basically at higher pulse frequency the reflection time is less than at lower pulse frequency.
This is a bit of a trade off, in that it does not make the engine produce as much power as it possibly can. Instead it flattens the torque curve getting rid of the notorious 2 stroke torque peak.
It is also worth noting that the primary reflection that comes back to the engine is from the flat plate at the end of the diffuser. (The right inner wall in the picture above.) The high pressure wave reflects off of the plate just like a ball bouncing on a flat surface. The amount of pressure in the belly will tend to shape the pulse, higher pressure in the belly makes a sharper and faster reflection. (When the pressure in the belly is low, the pulse dissipates to the sides more easily.) However ALL of the gas must eventually go back the other direction in the diffuser to the big holes as that is the only exit.

**This is intended to more closely match the torque curve to the torque curve that is required to properly drive a propeller.**

The above is the most important part that most land or water based 2 stroke vehicle mechanics and even some very knowledgeable engineers miss. What you will hear from the majority of them is that peak output is the primary goal... They all just talk about some added percentage of power by using a "real tuned pipe", or a "less restrictive pipe", among many other (wrong) statements, totally ignoring the fact that a normal tuned pipe works best at exactly one engine speed, any other speed it is out of (peak) resonance. It is a big part of the notorious sharp 2 stroke torque curve peak. It is this peak that they all base their numbers on. For a propeller, you need to spread this out and flatten it, which lowers the peak number. I hear the following more times than I can count, "The Rotax engines are junk engineering, I can get twice the power out of my XYZ engine that is the same/similar displacement." This simply makes the plane unflyable at any other setting than what produces peak power or within a small range of that. If you are flying aerobatics or competition, you might want a tuned pipe and maximum power! (I built a 145 HP 583 for this.) If you are doing anything else, you will definitely want the engine to be stable at lower power settings, so a broadband tuned exhaust is the only way to go. With a 2 stroke, it is critical to match the torque curve to the propeller, otherwise unless you have electronic control stability is impossible or at least extremely difficult to obtain at all throttle settings, and propeller/gearbox combinations. (Keep in mind that Rotax designs the engines to work in a large range of speeds and configurations, so they are not optimized for any particular one, nor does that even make sense to do so.)

Note: Rotax actually sells a muffler kit that is attached to the stinger. If you think that the can is quiet, you will be amazed if you add the muffler. ROTAX 503 | 582 UL 99/17 ENGINE AFTER-MUFFLER ASSEMBLY PARTS | California Power Systems
As you can see, they call it an after-muffler... This is something that has just caught on that is frowned upon by the Rotax engineers as it implies that the expansion chamber is a muffler as well. At one point in the late 80's when I went to the Rotax factory school in Austria, they were making a big deal to change the documentation to not leave the impression that the expansion chamber is a muffler as well, but at some point by the mid 90's when I went back, they seem to have given up on this. Also, they make it very clear that adding the after-muffler has no measurable effect on the operation of the engine or its power output.

Last edited:

#### Armilite

##### Well-Known Member
The can that many people call the muffler is actually not a "muffler". It is actually a broadband expansion chamber. If you were to remove it, the 447 is nearly impossible to get started, and it will produce about 20 HP give or take. Most people that are used to 2 stroke motorcycles do not recognize them as an expansion chamber for 2 reasons:
1) They are 50% the size of a single cylinder chamber, as they are designed to provide a reflection to the OPPOSITE cylinder via the Y pipe.
2) They are not designed to 'peak' the torque curve as are most land based vehicle 2 stroke pipes.

Some details:
Two-stroke exhaust pipes generally have specific sections:
2) Diffuser (divergent zone)
3) Belly
4) Baffle (convergent zone)
5) Stinger

Note: Some shops and engineers use different terms than the above and even have more sections defined. In the early 80's I wrote a software package for designing expansion chambers. I used different terms and more sections because of the way that the software did its calculations and derived reference points, (no other reason). As it turns out, this was a bad decision as it ended up causing many people to rename the parts, and I still hear the terms I defined in my software used in the real world at times... That was NOT my intention...

On the 447 and other similar Rotax engines, the header is the Y pipe and the elbow immediately after (if equipped). The diffuser is the tapered pipe between the coupling and the can. The belly is the center section of the can. The baffle is the tapered tube inside of the can AND the rear section of the can. The stinger is the pipe that exits the can. It is basically a folded design, so it is harder to recognize it as an actual expansion chamber.
Here is where things get really interesting. The big holes in the front end of the can in the sides of the diffuser are where a good portion of the actual tuning happens. These holes, both the size and quantity are critical, and it is in many models the only thing that changes between engine models. (IE the 447 and the 503). What happens here is that the faster the engine turns, the smaller the chamber appears to be to the traveling pulses. You can think of it as a cup with holes in the bottom and sides. If you pour water in slowly, it is mostly empty (large volume of available space inside). If you pour water in faster the cup will fill up more (less available space inside). What this effectively does is make the chamber appear smaller as the engine turns faster. Basically at higher pulse frequency the reflection time is less than at lower pulse frequency.
This is a bit of a trade off, in that it does not make the engine produce as much power as it possibly can. Instead it flattens the torque curve getting rid of the notorious 2 stroke torque peak.
It is also worth noting that the primary reflection that comes back to the engine is from the flat plate at the end of the diffuser. (The right inner wall in the picture above.) The high pressure wave reflects off of the plate just like a ball bouncing on a flat surface. The amount of pressure in the belly will tend to shape the pulse, higher pressure in the belly makes a sharper and faster reflection. (When the pressure in the belly is low, the pulse dissipates to the sides more easily.) However ALL of the gas must eventually go back the other direction in the diffuser to the big holes as that is the only exit.

**This is intended to more closely match the torque curve to the torque curve that is required to properly drive a propeller.**

The above is the most important part that most land or water based 2 stroke vehicle mechanics and even some very knowledgeable engineers miss. What you will hear from the majority of them is that peak output is the primary goal... They all just talk about some added percentage of power by using a "real tuned pipe", or a "less restrictive pipe", among many other (wrong) statements, totally ignoring the fact that a normal tuned pipe works best at exactly one engine speed, any other speed it is out of (peak) resonance. It is a big part of the notorious sharp 2 stroke torque curve peak. It is this peak that they all base their numbers on. For a propeller, you need to spread this out and flatten it, which lowers the peak number. I hear the following more times than I can count, "The Rotax engines are junk engineering, I can get twice the power out of my XYZ engine that is the same/similar displacement." This simply makes the plane unflyable at any other setting than what produces peak power or within a small range of that. If you are flying aerobatics or competition, you might want a tuned pipe and maximum power! (I built a 145 HP 583 for this.) If you are doing anything else, you will definitely want the engine to be stable at lower power settings, so a broadband tuned exhaust is the only way to go. With a 2 stroke, it is critical to match the torque curve to the propeller, otherwise unless you have electronic control stability is impossible or at least extremely difficult to obtain at all throttle settings, and propeller/gearbox combinations. (Keep in mind that Rotax designs the engines to work in a large range of speeds and configurations, so they are not optimized for any particular one, nor does that even make sense to do so.)

Note: Rotax actually sells a muffler kit that is attached to the stinger. If you think that the can is quiet, you will be amazed if you add the muffler. ROTAX 503 | 582 UL 99/17 ENGINE AFTER-MUFFLER ASSEMBLY PARTS | California Power Systems
As you can see, they call it an after-muffler... This is something that has just caught on that is frowned upon by the Rotax engineers as it implies that the expansion chamber is a muffler as well. At one point in the late 80's when I went to the Rotax factory school in Austria, they were making a big deal to change the documentation to not leave the impression that the expansion chamber is a muffler as well, but at some point by the mid 90's when I went back, they seem to have given up on this. Also, they make it very clear that adding the after-muffler has no measurable effect on the operation of the engine or its power output.
===============================

Rotax Exhaust are not Tuned Pipes, they are a combo Megaphone Header (Small to Large that runs into a Muffler. They also sell an add-on secondary Muffler or Sound Modulator. A 447 is 436cc and with a Sled CAN Muffler they make 40hp@6500rpm, so Rotax's UL Megaphone Header (Small to Large that runs into a Muffler doesn't Gain you any more Hp, still 40hp. If you cut off the Muffler and ran the Megaphone Hear it would Gain you some hp, but it would be very very Loud.

A True Tuned Pipe is made up of Sections as you say! It's Best to not Directly Bolt-On a Tuned Pipe but to use Springs for the vibration. When you Design the Tuned Pipe you need your Exhaust Port Width & Height & Lenght. The Exhaust Port Measurement for the length starts at the Center of the Port against the Piston Wall to the outside of the Cylinder + your Gasket Thickness + your Wye Pipe Centerline Length. That is where the Tuned Pipe Connects so you have the Proper Total Lenght. That is for a Straight Tuned Pipe!

Then you have to use a CONE Program to Slice these Sections up to make Bends!

2) Diffuser (divergent zone)
3) Belly
4) Baffle (convergent zone)
5) Stinger

You can get as complicated as you want to fit one.

#### Attachments

• 66.8 KB Views: 10
• 106.6 KB Views: 9
• 26.6 KB Views: 9
• 42.5 KB Views: 8

#### Armilite

I like these guys Tuned Pipe Software, Pipe Designer $40, and Cone Designer Free. #### Attachments • 76.5 KB Views: 5 • 47.7 KB Views: 4 #### n3puppy ##### Well-Known Member =============================== A 447 is 436cc and with a Sled CAN Muffler they make 40hp@6500rpm, so Rotax's UL Megaphone Header (Small to Large that runs into a Muffler doesn't Gain you any more Hp, still 40hp. Snowmobile 447 max hp is produced at 7000rpm not 6500 Same HP @ lower RPM, - the ultralight tuned exhaust is producing roughly 8% more power (torque) than the plain snow muffler. Last edited: #### koyama ##### Active Member You have to be careful when comparing between sled and aircraft engines. There were a few differing versions of the sled engines made that did not have anything obvious to tell which revision you were dealing with unless you are looking at the service manual. There are also many sets of specs floating around for the sled engines, some of them are from marketing, and they are intended to skew the numbers for sales competition reasons. With the aircraft version of the engines, the numbers were made to be specific to driving a propeller, and the marketing department was not allowed to mess with them. Also, other than at the very start of production the aircraft engines were built to the same specifications. (All of the internal parts that have anything to do with making power were unchanged.) The sled engine used at least 2 piston heights that I know of (the height of the piston is important as it sets the intake timing as the piston skirt is the intake valve.) They also used different carburetors and exhaust setups, as well as different ignition and timing. #### Armilite ##### Well-Known Member Snowmobile 447 max hp is produced at 7000rpm not 6500 Same HP @ lower RPM, - the ultralight tuned exhaust is producing roughly 8% more power (torque) than the plain snow muffler. View attachment 103467 ======================== "A 447 is 436cc and with a Sled CAN Muffler they make 40hp@6500rpm, so Rotax's UL Megaphone Header (Small to Large that runs into a Muffler doesn't Gain you any more Hp, still 40hp. " First, I didn't say the Sled 440F was Rated at 6500rpm, I said it Made 40hp@6500rpm. It actually makes more than 40hp @ 7000rpm depending on whether Single or Dual Carb, and the much later ones came with a Mild Tuned Pipe! The UL Engines didn't use Tuned Pipes. Only R&D Aero and one other company I forget the Name know made Tuned Pipes for Airplane use! A Stock 2004 380F was 36hp@7000rpm with Muffler with Dual 30mm Carbs, the 380HO was 48hp@7000rpm with Dual 30mm Carbs, and used a Mild Tuned Pipe with a 11.2cr vs the Standard 9.6cr. They never made a 440F HO to my knowledge! But if you built a 440F to 380HO Spec's it would probably be 53hp@7000rpm. The Stock 2004 380HO with just a better-Tuned Pipe, Dynoed 57.26hp@7000rpm. The 380HO was making about 53hp@6500rpm. FACT! If you want to get Technical, most 447UL's used a Single 36mm Carb! 447UL Single Carb was 39.6hp@6500rpm! Single 36mm Carbs! 447UL Dual Carb was 41.6hp@6500rpm! Dual 34mm Carbs! #### Attachments • 75.4 KB Views: 1 • 58.3 KB Views: 1 • 27 KB Views: 2 • 27.3 KB Views: 1 Last edited: #### n3puppy ##### Well-Known Member I wonder if anybody knows/understands the workings, and perhaps recipe for the rotax mufflers we see on nearly all 2 stroke rotaxes.. Cptcliffhanger - you have seen the very detailed description of the Rotax tuned exhaust from a Rotax factory trained and certified technician (Koyama) Here are some other explanations of how the Rotax tuned exhaust on your 447 improves performance. Mike Stratman, founder of California Power Systems, is another Rotax factory trained and certified technician. He wrote and published many technical papers on Rotax engines. Here is one about how the Rotax ultralight tuned exhaust works and the importance of exhaust tuning length. The "echo" timing depends on proper length. (Rotax dimensions given) https://www.cps-parts.com/cps/pdf/Part8.pdf This Stratman tech article expands on the first concerning Rotax exhaust performance. https://www.cps-parts.com/cps/pdf/Part48.pdf "Thus the length and taper of the exit cone is crucial. These designs are determined by extensive dyno testing by the Rotax factory under an actual prop load using variable length and shaped cones." There is also this article from Ultralight news - how changing the tuning length of the factory 277UL exhaust reduced thrust by 35% ROTAX 277 CC ultralight aircraft engine troubleshooting Last edited: #### koyama ##### Active Member I like these guys Tuned Pipe Software, Pipe Designer$40, and Cone Designer Free.
I am not sure this is even relevant. (I am not completely familiar with this software), however, it appears to be only for single cylinder. It does not appear to support Y pipes. As far as I can tell, unless you are building dual exhaust, this is not applicable, nor is it relevant to this conversation.

#### Armilite

##### Well-Known Member
I am not sure this is even relevant. (I am not completely familiar with this software), however, it appears to be only for single cylinder. It does not appear to support Y pipes. As far as I can tell, unless you are building dual exhaust, this is not applicable, nor is it relevant to this conversation.
==================================

Racers tend to use Single Tuned Pipes for Each Cylinder. How you attach that Tuned pipe all comes into play, Direct to Engine vs using a Stub Short Pipe with Springs which is best. The most Important Part is the Total Lenght of the Tuned Pipe! On a Twin when one Cylinder is at TDC the other is at BDC. So if you use a Wye Pipe when you do your Calculations, you have Exhaust Port Height, Exhaust Port Width, and Exhaust Port Lenght. That Exhaust Port Length is from the Center, Side of Piston to the end of the Stub Short Pipe, or end of the Wye Pipe, where on both the Tuned Pipe is then mounted. There all Designed for a Specific Max rpm, and Port Duration, and that's why they're not interchangeable with other engines. There is a Cheap Mans Formula for adapting Tuned Pipes. Like my 1971 Skidoo 335 20hp@5500rpm Single is (78mm x 70mm) 155 Port Duration and the Skidoo 670 Twin 115hp@7750rpm is also (78mm x 70mm) but 180 Port Duration. To use the 670 Tuned Pipe you have to add on (1) Inch. Since the 335 Can Muffler bolted on direct, no problem, use a Stub Short Pipe so you can attach it with Springs.

Here is a Stub Short Pipe for the Rotax 277. Too Long but you get my meaning. You may need to tweak it's Lenght Longer or Shorter to get it making the power where you want it for Plane use. A guy who uses a Sled 670 Tuned Pipe on his 277F for Racing turns it 9200rpm making around 55hp. Most Sled Tuned Pipes are Designed for 7000rpm Fan Cooled to 8000rpm Water Cooled. So adapting a Tuned Pipe that was designed for 7000rpm is going to be easier to make work for 6500rpm.

When Designing your Tuned Pipe you have to have an idea of what hp is possible at your Max rpm you want to turn it. When I talked to a Porting Specialist years ago about my 277UL, I told him I wanted to turn it 6500rpm, he said on Avg if using 11.5cr, a Proper Sized Carb, and turn it 6500rpm it takes on avg 7cc to make 1hp. The 277UL is 268.8cc/7cc= 38.4hp@6500rpm. Some People like to use % from known hp rating, like 26hp@6250rpm + 45% = 37.7hp@6250rpm, 28hp@6400rpm + 45% = 40.6hp@6400rpm! R&D Aero Dynoed a Stock 277UL and it made 25.4hp@6000rpm and with their Mild Tuned Pipe it made 30.5hp@6250rpm! 25.4hp to 30.5hp = 18.2469% difference. Some Engines use a lower CR, and Small Carb, that you have to deal with. A Good Engine is the Simonini Victor 1 Super 400cc rated 54hp@6500rpm using 9.5cr. 400cc/54hp= 7.4cc to make 1hp. 400/7cc= 57.1hp. But 11.5cr to 9.5cr = -2hp then becomes 55.1hp. The reason it's not making its full power potential is the Carb is (1) Size too Small. But still pretty darn good.

Stock 277UL vs 277UL with the R&D Aero Mild Tuned Pipe.

#### Attachments

• 19.3 KB Views: 1
• 110 KB Views: 1