Weils ja hier doch inzwischen einige gibt die unbedingt was rechnen wollen. Der Autor ist Neels van Niekerk (EngMod2T Software, http://www.rddreams.com/neels/engmod2t.pdf )
Quelle: http://forums.pstuning.com/index.php?s= ... #entry5894
Two-Stroke Engine Simulators
1. The Purpose of a Simulator
An engine simulator does not calculate the best tuned pipe (expansion box) or porting or inlet system or crankcase volume etc. for you. You decide what changes you want to make to the engine and then run it through the simulator to see what the effects will be. The accuracy of the results will depend on the accuracy of your input data and the sophistication of the models used in the simulator. It is entirely possible to input completely unrealistic data and have the simulator spit out results. It is up to the user to verify his data, both the input and output data.
The simulators are based on 1-dimensional gas dynamics. This means that in the ducts they use one dimensional flow with special models for gradual and abrupt area changes like diffusers, reverse cones, steps and restrictors/throttles. As we all know the flow in the ducts are highly turbulent and 3-dimensional. To solve for these flow cases is entirely possible but would require super computers and run times of days rather than hours or minutes.
The cylinders, crankcases, air boxes, boost bottles and exhaust boxes are solved as volumes with no flow calculations. Special models for scavenging and combustion are used. More advanced simulators allow the coupling to 3-dimensional CFD (Computational Fluid Dynamics) software to calculate the flow inside the volumes but this requires the availability of the CFD software (at a huge cost) and serious computing power with long run times.
Thus, the purpose of a simulator considering all the above is not to accurately predict the power output of an engine but to accurately predict the effect of changes to that engine without having to cut metal.
2. How simulators are ranked in the two-stroke world
A 2-stroke engine has one thing that sets it apart from its 4-stroke cousins and that is the effect of temperature. It has much larger effects on the 2-stroke engine performance as the engine depends to a large degree on the wave action in the pipes. The wave speed in the pipes is a direct function of the gas temperature. The higher the temperature the faster the wave speed. We use this effect to get good overrun performance - leaning the mixture and/or retarding the timing slows the combustion process down resulting in less expansion work in the engine and higher pipe temperature which in turn results in more overrun. It is important for the simulator to accurately capture this effect. This is done mostly with the combustion model. They are from simplistic to sophisticated as follows:
i) Burn rate, heat release characteristic, delay time and burn time (combustion duration) prescribed internal in the software
ii) The same as before but the burn time prescribed by the user.
iii) Burn rate, heat release characteristic, delay time and burn time (combustion duration) prescribed by the user. Using typically something like Vibe functions for the heat release characteristic.
iv) Burn rate, heat release characteristic, delay time and burn time (combustion duration) calculated by the software using a turbulent entrainment model based on some geometry, turbulent intensity, mixture purity and strength and temperature and pressure.
v) Full 3-D combustion calculation by connecting to a CFD package.
The first three, although increasing in sophistication fall flat when overrun is calculated except maybe the third one if the correct characteristics can be supplied. The fourth one, the turbulent entrainment model, fares a lot better but being a flow model still has some drawbacks. A typical scenario is to use the baseline simulation run to generate the parameters for model three and use that from there on. The fifth model is the best but outside the reach of normal users, requiring expensive yearly licensing and computers.
3. A brief discussion of available simulators
Top of the range is Virtual 2 Stroke of Optimum Power Technologies but at US$20000.00 plus per year it is out of the reach of most of us. Some board members has/had the opportunity to use it, mostly as students at some university. It can do multi-cylinder naturally aspirated, supercharged and turbocharged engines with carburetors, indirect or direct injection, with and without catalytic converters, piston, rotary valve or reedvalve inlets, etc. If you can think of something it can do it. It has optimization and distributed computing capabilities. It has a very flexible user interface. It is based on the work of Prof Blair but with some further development to the models by OPT.
In the same class as Virtual 2 Stroke is Boost by AVL, GTPower by Gamma Technologies, Wave by Ricardo and the Lotus engine simulator by Lotus Engineering. How good their 2 stroke capabilities are I have no idea.
Also based on Prof Blair’s work but for single cylinder use only is the Model No2 software available from SAE. It is very cumbersome to use. Bimotion wrote a nice preprocessor for it. It is meant more as a teaching aid with Prof Blair's book. It has a prescribed burn rate model type 3.
Mota6 is a combination of Prof Blair's models and more mainstream gasdynamic methods. I think it uses the 2 Step Lax-Wendroff method to solve pipe flows and an older version of Prof Blair’s scavenging model. It is a great piece of software that runs fast and allows a large number of iterations to be tested in a short space of time. I think its predictions are a bit optimistic. It is for single cylinder engines or multi cylinders where there is no connection between the cylinders thermodynamically or gasdynamic, only mechanically. It has a prescribed burn rate model of type 2.
EngMod2T is based on Prof Blair's latest published work and a few bits of unpublished work. (In around 1999 Prof Blair signed an agreement with OPT not to publish or disclose any further simulation models, as he was close to retirement and they pay him well it is probably understandable, profs don't earn all that much) The purpose of EngMod2T was initially for me to understand why certain things work and others don't. It uses more sophisticated models than Mota6, not necessary more accurate but capable of supplying more in-depth data. This sometimes requires more complex input data. This results in longer run times. It can do from 1 to 8 cylinder engines, inline, vee or flat configurations. The flat engines can be separate crankcases (flat) or common crankcases (Boxer) configuration. It has been used to model and develop a Mercury V6 powerboat engine. It has both the burn rate prescribed model of type 3 and the turbulent entrainment model of type 4 available to the user.
Dynomation2T I have only used as a demo model but did not like it much. It allowed the user too little freedom with combustion modeling and it used dated methods for the gasdynamics. Dynomation4T has just been through a major revamp, maybe the 2T version will follow? I think it uses a burn rate prescribed model of type 1.
Softengine I have only briefly looked at but have no comment for now.
TSR software is not simulators but a so called "expert system". It is based on a large database of measured engines. It is a good starting point to develop the inputs for the simulation packages.
Vannik
Hier sind einiger der o.g. Programme verlinkt:
http://adardaine.free.fr/articles.php?lng=fr&pg=227
So und jetzt gehe ich aber mal so was von in' Wald !
Simulations-Software
Moderatoren: solo, Kilroy, Stefan Philipp (M)
-
Rocksternchen
- Beiträge: 41
- Registriert: Di 9. Okt 2007, 19:02
- Wohnort: Sinsheim
-
fischergeiz
- Beiträge: 135
- Registriert: Mo 12. Mär 2007, 21:22
- Wohnort: Wesel (NRW)
Wow, der "Port map analyzer" ist ein verdammt interessantes Programm. Die lästige Rumrechnerei bzgl. TA´s scheinen wohl endlich vorbei zu sein.
Da stellt sich mir nur noch eine Frage:
Einzeln kostet jedes der beiden Programme $16,25, zusammen jedoch nur $23.75. Eigentlich ein super Preisangebot, wenn man die Einzelpreise betrachtet. Der Haken an der Sache ist nur, dass ich keinen Vorteil darin sehe die paar $ extra für den " Porting Calculator " zu bezahlen, da alle relevanten Features auch von dem "Port map analyzer" beherrscht werden.
Ist das wirklich der Fall oder übersehe ich da einfach irgendwas?!
(das Einzige, was mich an dem ersten Programm richtig interessiert, ist die "ring flutter"-Funktion)
Da stellt sich mir nur noch eine Frage:
Einzeln kostet jedes der beiden Programme $16,25, zusammen jedoch nur $23.75. Eigentlich ein super Preisangebot, wenn man die Einzelpreise betrachtet. Der Haken an der Sache ist nur, dass ich keinen Vorteil darin sehe die paar $ extra für den " Porting Calculator " zu bezahlen, da alle relevanten Features auch von dem "Port map analyzer" beherrscht werden.
Ist das wirklich der Fall oder übersehe ich da einfach irgendwas?!
(das Einzige, was mich an dem ersten Programm richtig interessiert, ist die "ring flutter"-Funktion)
