Vehicle Powertrain Modeling and Full Vehicle Simulation Using MATLAB and Simulink for Student Competitions

Hi, everyone. This is Veer from the MathWorks Student Competition team. And welcome you all to the second session of the Vehicle Design Webinar Series for student competitions. So today we'll be talking about vehicle powertrain modeling. We'll touch base on that. And then we have Krishna Teja from JK Tyres. So he'll be talking about full vehicle simulation. And he has a lot of content to share with you. So you have to be with me for some time.

And along with me, I have Sudhakar, Pratyay, and Himanshu who will be supporting the event. Now just a few recap things for the webinar. So this is the CDs which we are doing this October and the month of November. So first one we conducted on vehicle dynamics focusing on vehicle dynamics simulation. So if you have not attended it, you won't be getting the links for the recordings.

However, we'll be publishing the recordings. So if you haven't, you can reach out to us at racinglounge@mathworks.com with the email saying that you want the recording. Or you can also join our Facebook group, the link to which we'll be sharing at the end. Then today we have this completely dedicated session on vehicle powertrain modeling and full vehicle simulation.

And then on 11th of November, where my colleague Sam will be presenting. So it will be beyond simulation. You'll touch base on how to communicate with CAN networks using vehicle network toolbox. So if you haven't, again, we have shared the link in the chat. You can register for the 11th November session as well. There you will be learning communicating with the CAN network using vehicle network toolbox.

Now there are a few quick important things before we deep-dive into the webinar. If you have any technical questions, please use the Q&A section. And make sure that you are sending to all the panelists. And then chat box, chat box we'll try to use it for interaction for interacting with each other. So we'll be asking some questions, only a few not a lot of questions. So you can answer that in the chat box.

And then if you have some non-technical things, for example that can you zoom in, zoom out? It's not audible. The video is not proper. That, you can mention that in the chat box. Certain questions coming up that you're not getting the emails. So we haven't sent the email yet for the last session. We are in the process to publish the recorded video for the last session. We will send you within five to six business days.

So you'll be getting that. Otherwise, if you're not getting that as mentioned, you can reach out to us at racinglounge@mathworks.com. Again, we'll share the contact details and the link to the Facebook group at the end of the session. OK. So in the last session, we covered this. We covered these topics, racing line optimization where we saw one of the submissions from-- Jacob submitted his project for MathWorks Excellence in Innovation Program.

And this was also an introduction to MATLAB. And we also share a lot of resources from where you can download the code. Then Sudhakar, my colleague, he spoke about the equation based vehicle modeling using Simulink, where he demonstrated how equations can be converted into a block diagram representation. Further, we use vehicle dynamics blockset to model longitudinal and lateral vehicle dynamics.

And we also saw how you can add one driver, which was a Stanley driver to track certain parts. And then finally, we also covered in detail the multibody suspension system, which is a part of Simscape vehicle template. And we also demonstrated that from 2021b, you can also add a tire model to your multibody vehicle model. So this all, again, if you have missed it, don't worry. We'll be sending all the resources. And you can reach out to us on racinglounge@mathworks.com for more.

So with this, we'll have some exciting agenda and a little bit tight agenda today. So we'll be focusing on physical system, modeling of physical system. The whole session, I would say 1 and 1/2 hour session will be approximately, I think, 50 minutes will be dedicated towards modeling of physical systems, wherein we'll see a very basic battery electric model with motor cooling which will be based on the fundamental blocks.

And then we'll a little bit improve, or you can say add more details to a model where we'll be seeing how electric vehicle can be powered by BLDC motor. And then in powertrain blockset, that is also one of the toolbox that we, we have multiple vehicle reference applications which you can use for your different design purpose. And that will check out one electric vehicle reference application.

And then we'll see how Simscape, which is you can see equation-based modeling and powertrain blockset which is a data-driven modeling, how you can integrate these two. And then Krishna will come with his full vehicle simulation. And he has a lot to speak and a lot to show, and also he has a very nice introduction of himself. And then we'll share some resources. And then we can take some questions.

So with this, as we mentioned that most of the part, we will be focusing on modeling physical systems. So let's bring one image in front of you. And a very, very easy question for all of you, what is this like from the image from the physical system layout what you see? What can you think of this? What is this? You can write your answer in the chat box.

You have one one minute model, OK. RC circuit, OK. Equivalent circuit? Equivalent circuit, it's a circuit. Yeah, that's correct actually in electrical terms. But in terms of application-wise, what do you think? Which part is it for an electric vehicle? Battery modeling, OK. OK. OK. Single-cell model. Yeah, that's a correct answer, I would say. Yeah.

OK, Krishna is also answering. Thanks, Krishna. Yeah, so this is a battery or single-cell model. Now, the idea behind just to putting a rough diagram of this RC equivalent circuit or battery model, one-cell model is that this is what the physical system looks like, or physical system layout when we say. Similarly, you can also make one layout for a DC motor.

So this is LC or battery equivalent circuit. And to model this physical system layout or physical systems, we have a very interesting and exciting amazing tool called Simscape, which enables you to model physical systems. And what are these physical systems? Systems like thermal system, electrical system, mechanical systems, something that you can see, you can feel you can touch, and defined by using laws of physics.

And the one important thing is that, I mean, you must be seeing a lot of text on the slide. But the most important thing is basically it is just mimicking the physical system layout on a simulation platform, which is Simscape. So consider a DC motor. Now if you have to make a physical system layout of this, it will be having certain physical components. What are those components? Those components are resistors, inductors, electromechanical converters. And then for the load, you'll be having certain inertias.

So now when you're modeling that in Simscape-- so Simscape provides a certain good amount of physical blocks, wherein all the equations have been written. And then these each connections, for example over here if you see, these resistors, inductors, electromechanical converters, inertia, these are already having a set of equations. And these are interacted by energy exchange. So even the physical signal, what you see, it is having certain meaning over here.

So as I said, it is just a replica or mimicking the physical system layout on a simulation platform. As an example, here is a DC motor where you can see these elements. These elements are already having the equations. You don't have to derive it from scratch. You can connect it the way you connect in a real environment. And what you can further do is, for example, we are giving you a lot of different physical blocks which are there in Simscape library.

And if you want to customize it, for that you can use Simscape language. For example over here, you can see that we wanted a lossy ultra capacitor, a custom-made block. So for that, you can use Simscape language. And then what you can do-- so it's a long list, but in simple words since this is a webinar dedicated for powertrain modeling and basically full vehicle simulation, you can make an electric vehicle model. And you can perform optimization.

You can do component sizing and selection. And this is all code generatable if you're connecting it to the hardware or sending-- if you want the code to be done on a hardware. So important point from this is it is mimicking the physical system layout in a simulation platform which is Simscape. And then there's an energy transfer involved between these two components. Now once we say energy, this energy flow in this physical system, that is dependent upon across, and the true variable.

Now why are we telling this? Because this is helpful when you're making the connection. Because it might happen that if you want to send something like torque or maybe you want to send angular velocity, in that case how to do the connections. So across variables, I mean if you want to sense that, that is connected in parallel. And through, that is connected in series.

So here's one example in electrical terms. So power in electrical terms can be defined as voltage into current, where voltage is across, and current is a through variable. So when you want to measure or sense voltage, that has to be connected in parallel. And if you have to measure current, that has to be connected in series. Similarly, in mechanical terms, this power can be defined as angular velocity and torque. And now if you're making the connection, this is how it will look like.

So in the first top diagram, you can see that we want to sense the angular velocity. So these are connected in parallel. And then we want to sense the torque. For that, we are putting the sensor in series. So these are certain basic things, and I mean, a lot of other things are also there. Eventually when we move through step by step, different models, you'll get an idea. And we'll also share you some resources. But you must be wondering that how do I remember all these? I mean, which variable is across, which variable is through?

So for that, you don't need to remember. You can just follow the sheet. This is also in the documentation. We'll be sharing that link as well. And even when we start the modeling, at least if you have any confusion-- for example, isothermal liquid, you can see that across variable, it's absolute pressure. And then through variable, it's a mass flow rate. So you can just follow this, and you can build your own model.

Now what we'll do next is, well, we started with this equivalent circuit model. So we are not going deep-dive into battery modeling. But then just this motivation of this problem statement, all this physical system layout, we'll see how connections are made and how easy it is to make the connections and few important things, what we have got. Once we are done with this, then we'll move to a live demo of electric vehicle-- basic electric model with motor cooling using fundamental blocks. So let's switch to MATLAB.

So let me go through this first, the library browser where you can find these blocks. So once you open the library browser, you can see over here that-- as you scroll down, here you can find Simscape. And in Simscape, you have drive line. You have electrical, fluids, fundamental library, and multibody. So in the last session, we spoke about multibody and a few-- also we've touched base on driveline.

So since we are looking into an RC equivalent circuit, or a battery current circuit, let's get into the fundamental library and see where you can find all those physical blocks or elements. So in the electrical, we will go inside electrical elements. You can see over here that you have these electrical elements, capacitor. You have resistor. You have inductive, which are all needed for when you're going for these kinds of circuits.

So now let me do one thing. I'll just add the image in this blank canvas. So if I go to this. So this is just a way to easily build a new model. For example, if you are planning to make any model, for example, this battery or DC motor modeling, in that case, this is how the first step should be. If you have a layout of the system or equivalent circuit, whether it is electrical or even in mechanical, you can take the example of spring mass damper system.

So just keep an image. And you can have this and the library on one side. And you can just keep on just adding the different elements or physical blocks over here. So for example, I just dragged in this capacitor. And when you double-click on this, you don't have to define any equation over here. Over here, what you need to do is that you need to define the capacitance and different variables. And one more important thing is that whenever, whether you use any mathematical modeling or physical modeling approach, the important thing is to get the realistic results, how you define the parameters are really important.

So we have seen students working on this physical modeling. They have a nice model with all the proper connections. But if you don't feed the correct parameters, then it's hard to get proper results. And this makes sense also for any kind of modeling method. So we'll bring in capacitor. Similarly, we'll bring resistor. And you can see we'll just make similar connections over here. So like this, and we'll just bring one voltage source. And we'll just simply connect it.

So you can see that we're just mimicking or replicating the physical system layout over here. And then you need to just connect it to the load. I mean, we'll see how we are connecting to different loads and full vehicle So this is, I mean say just an example of how easy it is to just mimic the physical system layout over here. Yeah. So again, what we'll do is, I mean we'll take this as a motivation.

And when we are moving to an electric vehicle model, we'll show you how just with this equivalent circuit approach, we have built a whole battery pack, and then how we are simulating it to a vehicle dynamics model, OK? So this is just starting. Now I know that a lot of you are interested or looking forward to learning battery modeling. So fortunately, we have a good amount of resource. And the best resource for you all is this page.

You can visit this page, Battery Modeling. And you can get here a good amount of information-- information like battery modeling, finding the state of charge, state of health, and even battery management system. So you have videos, tutorials, documentation, and doc examples. You can have a look into that. OK. Now moving to our main goal, that is modeling an electric vehicle model, we'll start with a very basic fundamental model using fundamental library. Like for example, when I'm talking about the fundamental blocks.

So basically, these are all the blocks which are coming from here, the fundamental library. So now let's move ahead and see what's the problem statement? So consider this vehicle. And this vehicle is comprising a battery and is having a motor. Then it is having gearbox. And then it is driving the wheel. And when it is driving, it is having the aerodynamics impact.

Now if you have this model, what all can you do? First thing, you can size the components. You can estimate the battery performance. You can study the vehicle performance. Or you can also design a motor cooling system. So we'll be taking this as an example. And the scenario is something like this, where the vehicle aim is to overtake this. And from here, we'll derive or you will size the motor based on this scenario. And further, we'll also look into how you can build a basic, very simple cooling system for the motor.

And so what do you need to do is, for example, if you are starting with Simscape, either you can just start with a blank canvas, or you can just type ssc_new and it'll open one canvas with some important default blocks which will be, by default, needed when you're modeling it. OK, so we'll come back to what these are. So you just saw the vehicle was having one wheel, which was rotating, then it was having some aerodynamic effect. And it was also having a gearbox.

So what we'll do is we'll try to make this with a very simple approach. So first thing, what we need to do is we need to rotate that wheel, right? So for that, we'll just bring one idle torque sensor, torque source, where C and R are the mechanical conserving ports. And then we have a tire, a simple tire, where again A and H are mechanical conserving ports. Now to vary the torque, what we'll do is we'll add one gearbox.

So we'll add this to one pole with idle torque source. And then this will be connected. The other port of the gear will be connected to the tire. Now to define the inertia, to include the inertia, we'll just bring one mass. We'll make this two-port, because you have to make some other connections as well. We'll just rotate it and make the connection. So I mean, as we are building it, you can see that we are just replicating the physical world.

We want to rotate the gearbox through which it will rotate the tire. And then we want to include a mass, which is a vehicle. So a few things, we'll just attach this to a rotational mechanical reference. And then we have to add one constant over here, I mean, just adding some constant torque. Not the signal, but this constant. So now if you see over here, there's a change. What

Is the change? That if I type constant, by default it gives you one Simulink block. However, as we've already mentioned that, these are all built using physical signals. So that's why this is a PS constant, physical signal constant, which is connected to this ideal torque source. OK. Now further, what we want to measure, we want to measure the velocity of the vehicle. So for that, we just bring one idle-- OK, so we bring this idle translation motion sensor. Why? Because we want to measure the velocity.

And then we again do the same thing. We connect the conserving ports. And further, there's one more important thing, which is lying over here. That is a solo configuration. So as already mentioned that these are all physical signals, for this we have to add this solo configuration block in all of the physical networks. So what it will do is it'll make sure that it is solving this physical signal. And then overall, the model will be solve using the global software using Simulink model presented over there.

So further, there are further-- a little bit more connections over here, which will make it. And then we covered, I mean, how the connections are made. We covered why this solo configuration is present over here. And then another aspect over here is that already mentioned that these are all physical signals. So for that, if you want to visualize this with respect to time, you need to connect it to the scope. So for that we need to convert the physical signal into Simulink.

So this is how the architecture looks like. Now what we'll do is we already have the model with all the parameters because if I, again, start adding the parameters, it'll take a good amount of time. So I'll just close this one. And we will move back to the model, which is already there. OK. So now over here, you can see it's the same thing. We have this torque source. You have this gearbox, the chassis and everything. Now let's run this model.

So if you run this-- OK. Yeah, so if you run this, you can see that the velocity is continuously increasing. And one more thing, what, I mean, I forgot to mention is that when you're connecting this Simulink block to Simscape using these converters, here you need to define these output signal units. For example, this, we have defined kilometers per hour. And you can see that this is way too fast. I mean, you won't be able to drive with this one, this velocity.

So now what is missing? We are missing with rolling resistance. We are missing the inclination effect in the vehicle, and even the drag. So what we need to do is, again over here, we'll add something, some more blocks, one, two blocks to include that resistance. So going to that model. So now with that improvement, what we did is that we have added a block called translation friction, which is accounting for the track.

And then for the inclination, we have added one ideal force source to which we are giving some constant value. OK, now let's see what happens. OK, one second. OK. So now you can see that the effect is much better. It is after 400 seconds. You can see that the vehicle is approaching the constant velocity. What do you think? Is it fine, or we need to make some changes? You can type that in the chat box.

So if you're looking from, I mean, the send in the model and even the results, what you are getting, what do you think? What could be wrong in this model? Or is everything fine? Gearshift effect? OK, Aditya says gearshift effect. OK. So what is the effect of suspension? Yeah, I mean, coming back to the suspension, we have not added suspension. But still like the basic modeling, you can add a few more things.

Or the major concern over here is this part. You can see that the vehicle is moving very fast. So what we'll do is if you remember the scenario which we are talking about, we are talking about the scenario where the vehicle has to overtake the other vehicle in one lane and come back to its original lane. So what we'll do is here, we'll try to control the speed. So for that we'll add one driver-- the driver, which is dedicated for that kind of overtaking.

So going to that model to the driver, over here if you see-- now just so that you don't confuse, this driver model is not a part of any toolbox or not a part of any Simscape block. It is just created using Simulink blocks. So if we go inside this, and the task is to controlling the velocity. Now this driver model is having currently you can say that 200 Newton meter torque. Now let me run this first.

And now you can see that, OK to some extent it is fine. I mean, it is having 100 kilometers per hour over here. And then it is increasing the velocity when it is overtaking it. And then it is coming back to constant. OK. Now coming back to the component sizing and selection, which still is a question is that if you want to select a proper component or motor for this. So for that you have to get into or visualize the torque. What is the torque output?

So far that, you can use the Simscape result explorer, where you can explore the results from different Simscape blocks with respect to time. So if I go to gearbox, and if I see over here, and let's expand it. OK, I closed that window, but OK, let me again start that. So over here, you can see that it is taking around close to 15 seconds or 14 seconds to overtake, which is a good amount of time. Now to this, so first problem was controlling the velocity, which we did just adding one driver.

And then if you want to reduce the time to overtake that vehicle, let's do one thing. Let's increase this torque. So if I increase this torque and further than the model, and then again visualize the results, let's see what happens. So now we can see here that this is much better. Now it is taking around less than four seconds, which is perfectly fine. Now the whole conclusion of the story is that, I mean, when you have certain targets to achieve, based on that, you can build a model. And then you can size the components.

So from here, we can at least say-- I mean, there could be a lot of other parameters like I'm checking, seeing the chart. I mean, there's drag. People are telling the lateral dynamics. There are a lot going on. But the overall thing is like, this is how the procedure is like when you are looking for component sizing and selection. Now with this, at least we understand that if you're selecting some motor, which is having a maximum torque of 800 Newton meter, that motor, we have to select.

So that's what we did. What we did is in the next model we have added fan motor. So wherein you need to define certain parameters. This motor is designed based on velocity and torque. And then you have to define these certain parameters. And once you have defined it, and then when you run it, you can see that it is within the range. And it is below five seconds. And it is fulfilling all the requirements.

So we saw that we started with a very simple basic model. It is still a simple basic model. And how we decided on what should be the specs for the motor selection? So this is one part of the model. Now another-- OK, there are a lot of questions. OK, one I can take. Why connect driver between scope and PS converter? Why connect driver between scope and PS converter? So the thing is this question is asked I think two times, why connect driver between scope and PS converter?

The reason is that this is completely built in Simulink. So that's why what we're doing is we want to input the Simulink signal, which can be done using this converter. And again, we want to give back the motor over here, what you see. This requires the torque as a physical signal. So that's why we are using another converter here from Simulink to physical signal con-- that's why we are connecting it. Yeah. So I mean, you can definitely keep on asking the questions. Once we are done with the session, we can come back and go through it.

So having said that, we just saw the model. And now another task which we already spoke about was how to add the thermal effect, or how to design a cooling system. Now at least even if you are new, you got an idea that it's just that what you see in a physical environment, real environment, that's what we are trying to replicate over here. So with that learning, what do you think that-- I mean, and one more hint I'm giving you over here that this is this H port, what you see on this, this is the thermal port over here.

So when you double-click on this, you need to define certain thermal mass and initial temperature. So now suppose we want to add one cooling system, so what do you think? What are the components? What do you need? Or maybe before that, what kind of different heat transfer will take place from the motor to the environment? What are the different modes of heat transfer which will take place? If you have that answer-- I mean, I think it's very easy to understand when you're modeling the physical systems.

OK. Mayur says convection. Convection. OK. OK. So there's one interesting answer, conduction and convection. Conduction, OK yeah. Conduction and convection, nice. Fine. I mean, just on a simpler note, yeah I mean I think most of you are correct. Of course, convection will be there because it has to dissipate the heat into the environment. So it's a through fluid state of liquid convection. So that's one. And then conduction will be also there because motor will be in certain casing.

So for that, it is a heat transfer through solid medium. So it will be conduction. So it will be having-- so now when you are modeling it, when you are adding that thermal effect, it will be having conduction, as well as convection. So with that, we'll go to the final model, how it looks like. So currently just focus on this one. I mean, yeah. Yeah. Let me see. Yeah. So what we did, so as you all answered that connection-- yeah, this is with the casing.

It transferred through solid medium, and then to the environment that is heat convection. That's why we have put it over. And then we have certain temperature sensors and other source blocks. Now further to cool it, so the cooling how you can do the cooling, you can have-- I mean, here you are talking about liquid cooling. So that's what we have done is that we have added two pipes. It is a pipe casing which is connected to the casing over here, through which the fluid will be going.

And then to dissipate that heat, we are connecting that to the ambient or the ambient pressure, or you can see it, the atmosphere. And then there are certain blocks for which you have tank, pump. And then to define the fluid property, we have one more physical block. Now once you run the simulation, what you see over here-- Ahmad, can you send the model? Yes, we'll send the model. We'll show you something interesting, where you can find the model.

Now what we did is I just switched off the cooling system. And now let's check out the temperature, what's happening. So if you have some threshold value, something like 120, you can see that it is crossing. And the temperature of this motor is increasing as the time is increasing. Now when you add this cooling system, let's see what happens. So yeah, there's one question meanwhile I can answer is what Simscape libraries are needed for the model with cooling?

So yeah, I mean along with what you saw, we took some models. Made some blocks using fundamental libraries. But otherwise, you need to go into the fluids. And you can check out these blocks which you can find these pipes on. So yeah, I mean now I switched on that cooling. And you can visualize over here that this is within-- OK, not this. So now you can see that this is below the threshold point. And we are able to cool the motor.

So this is just, I mean I would say, very on a basic level. We showed the model. And now what we'll do is we'll move into, I mean, more detailed electric vehicle model, where you will see how we are adding a specific motor and battery as well. OK. I hope, I mean, you got some idea. Again, to download these models, this is the link. So this is basically what I spoke about, I mean what I demonstrated.

It is completely in the form of one of the videos so that you can find-- I mean, the models you can find over here, the battery electric vehicle with motor cooling in Simscape on the File Exchange, this is also connected to GitHub. And there's also a link to the video tutorial to this. So you can have the model from here. And you can also, again, watch the video to learn the same messaging basically. OK, so moving to the model, let me close a few ones.

Yeah. So we'll just quickly run one more model which is having more detailing. Because when you talk about the electric vehicle model, you have DC-DC converters. You have motor, specific motors. So now you can see over here that when we go to the vehicle block, we are not modeling it using mass and drag using a friction block. Rather, we have a dedicated vehicle body block, wherein if you double-click on this, you can see that this block is incorporating the equations, which is having mass, the number of wheels per axle, and then CG height and everything.

Now it also takes into account the drag, pitch, and all. So this is one block, and then it has certain connections for making connections with the tire blocks, and then brake as well, and putting an inclination. Now with this recent release, we have this interesting one more block in the driveline, which is called longitudinal vehicle, which is also incorporating the tire effect. So you don't need to separately connect tires like what you saw in this model. So you can also have a look into this one.

And further, if I go ahead, you can see gear systems which are modeled using these friction clutches. And based on that, we have certain logics. And then as we mentioned the battery pack, so if you go over here, you can see how we have utilized an equivalent circuit model to make one cell. And then we have connected that to make a whole battery pack. OK, just to confirm, is my screen visible to everyone? OK, so you can see the screen? OK fine. OK. Perfect.

So we're talking about the battery model, how we utilize a single cell equivalent circuit into a whole battery pack. And then coming back to the motors, now we are not using the motor, which is built on using torque and velocity values. Rather, we have a BLDC motor. Similarly, you can have PMSM motors. Also, you can check the Simscape library, and you can see what other motors we are having.

And then we have also built one DC-DC convertor to step down, and then three-phase inverter for this one. We're controlling switching pattern using computational logic. So again, you can see. So you can see as a learning perspective, you can start with a basic model. And then as you proceed further and as you understand the concept of whole physical modeling in Simscape, then you can add more details into the model.

Now how this model can be used? For example, in Formula 1 competition for component sizing and selection, if you are sizing your battery, if you are selecting your motors, I mean you must be having some kind of drive cycle. You can have that drive cycle over here. Add one controller, which could be simple or PI controller. And then you can see that, OK, the powertrain which are designed that is able to track the velocity or not.

Similarly, you can also optimize it. For example, you can run certain optimization to find the parameters of this motor. So that's what we wanted to show over here, how you can increase the fidelity. And if you want more high fidelity models, what we would recommend is do check out this resource, that Simscape vehicle template. We also spoke about this. And we also took one model from this and demonstrated during the last session, the suspension model for the formula student teams.

And you can get a lot of information from this even because it includes one-axle, two-axle, multi-axle vehicle models, and with the powertrain and vehicle dynamics as well. So you can explore this. We have provided the link over here. And feel free to use these models for your specific requirements. Perfect. So quickly, what we will do is we don't have much time, again, since we have to go to Krishna.

Until now, we just spoke about the concept of physical modeling. We also have one more tool called powertrain blockset with a goal providing starting point for engineers to build good plant/control models and provide open and documented models. For example, this toolbox has a good amount of reference applications. You can use that reference application by changing variables. Or if you want to make any changes to any of the blocks, these all blocks are open. And of course, those who are interested in or looking for HIL testing, it provides very fast-running models for that.

Now what is this powertrain blockset? It provides you different libraries for drivetrain, energy storage, propulsion, transmission, vehicle dynamics, and vehicle scenario builders. And you can combine these models to, for example, build one engine model. And then that subsystem can be added to a complete full vehicle simulation. And then you can visualize the results or do the analysis part.

So these are the whole set of reference applications over here, and this full vehicle model. And we'll just check out one example for the electric vehicle one. And then if you are looking for a virtual engine dynamo meter testing, you can also check out these reference applications on this first CI engine, and another one is for SI engine. So I'll quickly just open this model and show you how you can leverage this model for your specific requirement.

Yeah, so this is the model, what you see over here is it is having passenger car. It is having different subsystems. And if you go to the passenger car, and you go to the electrical plant, you can see how the power powertrain has been built using blocks from powertrain blocks. For example, if I just take example of battery, this is the block which is present in the powertrain blockset. Again over here, these parameters, you need to define to model the battery.

And just to, again, remind you, these are not physical signals. These are based on Simulink. These are based on equations. And you can see that these are having directions. So you can see input signals and output signals. Now if you want to change it or you want to visualize it, inside what it is, or you want to customize it based on the existing blocks, you can simply go inside this. Press Control-U. And you can see what the equations are being used over here.

So it's completely open. You can use it. You can modify it. Or you can directly just use the reference application, similarly with the other components as well. So now this model if we run, it will take some time. And since we don't have much time, just quickly I'll mention that this is a closed loop model. What it is doing, it is having a reference drive cycle. Again, this is the FTP75 drive cycle for this. We are running the model.

And then you can see how smoothly. It is tracking. And then you can also calculate other parameters like state of charge for this electric vehicle model. And we often get a lot of questions regarding, can we have other drive cycles? So if you just click on this Install additional drive cycle, it will route you to one of the software packages from our submission on the File Exchange. From there, you can get more drive cycles for testing. So with this, we'll just give it to Krishna. OK, so Krishna just to make you as a presenter let me know if you are able to get that access.

Thanks Veer for, first of all, inviting me for this webinar, and I hope-- it has been a great presentation already. And a lot of things and very detailed are already covered. So what I'll do is I'll first run you through a few slides which actually describe how a full vehicle can be modeled. But before modeling that, I would like to give a small use case that we have. And of course, it's a very big project, but I'll try to compress it within four or five slides basically to make you understand how different parameters affect, and in real vehicles how testing is being done on track.

So with that small intro, let's get started. So basically, I have done my PhD at IIT Madras, in terms of vehicle dynamics and reinforcement learning control. And Professor Krishna Kumar, I think many of you might be knowing his NPTEL and YouTube lectures on vehicle dynamics and finite element analyze. So I joined JK Tyres recently, and primarily focus on driver vehicle interaction, and basically using software-in-loop and driver-in-loop, hardware-in-loop simulators, tire tuning, and ADAS systems as well.

So just to give you a feel, this is a view of our lab. And at the bottom, you can see me driving the hardware-in-loop, or basically the driver-in-loop system as well. And the car that you see behind is the same, basically. But this is, of course, extensive parameterization, extensive modeling and all that stuff. And these are some of the things that we are going to explore in the forthcoming steps. In a few minutes, we're going to see that.

So Veer has already given you a very good basic understanding of the difference between equation-based, that's on the left side, and physical model, that's in the center. But the physical modeling is basically, again, categorized as one 1D and 2D. Plus in 1D, basically we do not have gravity or any other inertial effects, and so on. So that is where we go into 3D to model even more details and the combination with 1D simulations as well. So this is what we're going to explore more in detail.

So the simulation model basically has all these subsystems that you see that from the big chassis, suspension tires, brakes, steering, maybe power steering as well. everything is electronic power steering. Some, if you go to German vehicles, European vehicles, basically they do not have a mechanical steering. That's why it is called steer by wire. So everything is replaced by wires, so how those can be modeled. And powertrain is already covered. So I just will go to the basics.

And Formula students might be interested in aerodynamics. So I'll just give you a brief of, how do you model aerodynamics? Extensive modelings can be done, but here I'm going to touch the basics. So to give you a feel, I'll first run you through a case study. So the case study is regarding objective handling performance. So handling, it is basically extensively dependent on these subsystems-- tires, suspension, and steering system, of course. Of course vehicle level parameters and other parameters are not very well felt, but these are the three main parameters.

So for these performances, for example a new tire is going into the market, or a new vehicle is going into the market, it needs an extensive validation not just on the sensor side, but also the physical feel of the driver. That is, what is the subjective rating needs which the driver himself gives? So we require different metrics from individual components. We require overall metrics of the vehicle itself.

And the challenges are pretty big. We need a full vehicle CAE simulation and overall process efficiency in data collection, modeling, simulation, testing, so on. The biggest challenge, which physics to model, and which we can permit. So the trade-offs are huge. So these are the things that we need to address before. And once these are done, we can actually have a correspondence map. Which performance metrics is affected by which parameters? So that is what we are even going to see in the course of this presentation.

So first, we are going to measure the tires. So on the left is what you see as called a construction trailer or skid trailer, basically, which measures all the tire force and moments under different conditions, basically. And from those raw data, this is a skid trailer that is available needed. It is the first that we have purchased. And it is used extensively for tire measurements.

Next, we estimate all the tire data, again, using MATLAB extensible optimization and estimation techniques, codes built-in so we can directly leverage them. Finally, we will perform the simulation, full CE simulations as you can see on the right side. Next is vehicle parameter estimation. Of course, you can have data from KNC machines, or we have exploitation protocols. Some of them are published online. You can get these papers online.

So once we model them and use the parametric estimation with sensors that are mounted, you can see here, a few sensors that are used for suspension. And for vehicle level testing, we have RT 3000 steering robotics on which to capture and maintain certain conditions of test protocols. On the right side, you see the measured data and estimated data much better.