Why and what of Autonomous HVAC CFD

We have recently launched our new product Autonomous HVAC CFD. On this launch, I thought of telling you the story. A story that talks about the “Why and what” of the product. This is little behind the curtain stories, little philosophical, and little about the vision. I hope you will enjoy reading and knowing more about this new star across the block.

What is “Autonomous HVAC CFD” ?

First thing first….deciphering the name

So, the name of the product is "Autonomous" "HVAC" "CFD". Internally we call this product AHC, and we wish you all remember this sweet little acronym. Let’s see what exactly the name stands for and what is the philosophy behind it.

HVAC

This stands for Heating Ventilation and Air Conditioning. This refers to different systems used for moving cold or hot air between indoor and outdoor areas. These are the systems that are supposed to keep us warm and cool depending on the conditions surrounding us.

HVAC is a huge domain and there are a lot of activities involved when it comes to designing this system (Yes, you guessed it right, the product is something to do with designing HVAC systems.). AHC product is focusing on a particular aspect of HVAC system design. That is “How do we know if a designed HVAC system provides required thermal comfort for occupants?”. The product starts with answering this question by evaluating the thermal comfort of a given HVAC system design. We feel that there is a bigger question to be answered. The question is “May I know the performance of multiple HVAC systems I have designed so that I can compare them and select an optimum one satisfying the occupant thermal comfort need”. AHC product is mainly conceptualized around this question.

The little secondary questions around these are:

Although these are secondary questions in the context of the design problem, they are very important considering the demanding market we have around. So, the core of the AHC product is solving the primary question, but the product is beautifully crafted to give answers to all the secondary questions too. So, to summarize "HVAC" part in the name, AHC product is

A product which quickly evaluates the thermal comfort performance of multiple HVAC system designs

CFD

This stands for Computational Fluid Dynamics popularly known as CFD. It is a technique to solve fluid and heat transfer physics. Flow and heat transfer physics is well understood, and a set of mathematical equations are developed to represent this physics. These equations are known as “Navier-Stokes equations”. If we solve these sets of equations, we will be able to get the values of different flow parameters in the region we solved these equations for. Although it looks simple and straight, these equations are very difficult to solve especially in complex regions like in-door sitting arrangements with thermal equipment.

Many researchers have worked on this problem and came with a mathematical method to solve these complex equations. The solution is established using some numerical methods to solve the equations. Let’s not get more into these methods for now but let’s just say that CFD is “a numerical method to solve fluid and heat transfer equations”.

Estimating the thermal comfort of a given HVAC system requires calculating the velocity, temperature, and humidity at various locations inside the domain. Once these basic parameters are known, we can estimate thermal comfort parameters like Predicted Mean Vote (PMV), Percentage of People Dissatisfied (PPD), Effective Draft Temperature (EDT), Draft Rating (DR), etc. In the AHC product, once the user defines the supply conditions, we use CFD to calculate flow and thermal parameters. The primary variables like velocity, temperature, and humidity are calculated at various points in the region. These values are then used to estimate the overall performance parameters of the HVAC system. So, to summarize the "HVAC" part in the name, the AHC product is:

A product which uses CFD technique to calculate occupant thermal comfort

Autonomous

The meaning of the word autonomous is contextual. It depends on the industry we are referring to, the device we are talking about, and the problem in which an autonomous solution is implemented. A few of the general definitions for autonomous are :

“A system which operates without human intervention.”

“An autonomous system is one that can achieve a given set of goals in a changing environment – gathering information about the environment and working for an extended period without human control or intervention.”

“An autonomous system is the one in which hardware and software work together to solve a problem by performing an action.”

At CCTech (the parent company of simulationHub and its various products), our mission is “Transforming human life by the democratization of technology”. This being a guiding pillar, the simulationHub team works on developing various products supporting this mission. We believe that irrespective of how complex the technology we use in the back-end, all our solutions should be “autonomous” enough so that end-users should be able to harness the power of that technology. The first “autonomous” solution we launched was for the valve industry. Our "Autonomous Valve CFD" application is now being used by various valve designers across the globe. The product is used to conduct 1000s of valve design analyses each month.

When we started conceptualizing the solution for the HVAC industry, we knew that we must develop an “autonomous” solution irrespective of the challenges posed, and irrespective of the complexity of the problem. We knew that we should develop a product that should drive the solution without much intervention and inputs from the users. The product should be intelligent enough to adjust the process to get the end solution. With this BHAG (Big Hairy Audacious Goal), we set out on the joinery to develop the product. It took almost 2 years of the rigorous software development cycle to come up with the autonomous solution we were looking for.

Except for a few user inputs, the product does everything autonomously without asking a single CFD-related question to the user. All the CFD decisions like “selecting the mesh”, “selecting appropriate boundary conditions”, “selecting the physics to be solved”, “extracting the required data from solution” are done without user intervention. All these decisions are critical, complex, and driven by system design. For products to drive the solution autonomously, we have developed various intelligent algorithms and used many machine learning techniques. If you have used CFD in the past or know a little bit about it, you will feel like there is some magic in the product. But, believe me, it's not the magic, but the intelligence available in the product which keeps listening to your inputs and adjusting its behavior. So, to summarize the "Autonomous" part in the name, the AHC product is:

A product that keeps listing the user inputs, adjusts its behavior, and generates the solution autonomously

I hope that you get the gist of all the parts in the name of the product. So let me now take the liberty to provide the full decoding of the product name. Autonomous HVAC CFD (also known as AHC), is :

An adaptive, and autonomous product, which quickly evaluates occupant thermal comfort of multiple HVAC system designs using CFD technique

In the rest of the article, I am going to talk about various components and features of the product. The focus is not on what the feature is, but why we developed that feature in a particular way and why it is part of the complete product.

BIM designer

We have been working with the HVAC industry for more than 10 years. Over the years, we have understood their design process and the products they are using at various stages of the process. For designing and communicating the HVAC system layout, designers mainly rely on products like Autodesk AutoCAD. Designers also use products like Autodesk Revit to create the 3D model of the HVAC system design.

Creating a CFD-ready model from these design tools has always been a big challenge. Designers either need to recreate the complete model or do a lot of simplifications. During the first phase of the AHC development, we worked on various options to take the models from these design tools and generate CFD-ready models. The majority of the research was around Autodesk Revit and how we can import the model as is. After a few months of research, we realized that supporting such functionality is going to take a lot of development efforts. Although this feature is in our development road map, we thought of pushing this to future product releases.

But the basic input needed for the solution is a 3D model of the system. After a few more weeks of research, we came up with the concept of using a simple “sketching” tool to generate the indoor layout. This idea led to what is known as “BIM designer” in AHC. This intelligent BIM designer is used to create the built environment quickly. You just keep creating the layouts by sketching the walls, and all required CFD information is generated automatically. The BIM designer also has library components like supply, return, windows, furniture. Users can drag and drop these components, use default properties, define custom properties, and generate the HVAC system layout they want.

BIM designer

This is one of the major components of the AHC product. It helps in the complete autonomous solution journey by :

Library components

Accurate prediction of occupant thermal comfort needs accurate values for all heating and cooling requirements. These requirements are governed by the heat load inside a space, wall material, window configurations, etc. Many times, these accurate values of head load, properties of walls and windows are not handy. The HVAC industry has developed many air distribution products. Each product has its unique design and performance. It is difficult to predict thermal comfort without capturing the correct flow pattern coming out of these distribution products.

Considering this, the major challenge in AHC development was to make sure that the user provides accurate heat load, supply temperature, and flow rate data. We thought of many options like excel based inputs, allowing importing of standard library data, or asking users to build these components from scratch. But we wanted this setup to be as quick as possible. We wanted to have a component that identifies the design intent and generates required data. We wanted to have components with default values and flexible enough for users to modify as per the need.

So, we came up with the concept of an in-built library. This does only have component information but also contains a 3D model of the component. The user just needs to select the component, drag, and drop it in the room layout created. Once placed, users can move, rotate, or edit component parameters.

Library components

All the library components have default heat load and flow properties. These properties are based on our extensive research and understanding of the various elements involved in HVAC system design. These default values will satisfy about 90% of the system design needs. If needed, users can also edit the default values.

The library components help designers to build the configuration quickly. These intelligent components provide all the inputs required for the autonomous generation of CFD models.

Design configuration and scenario

This component has probably gone through most design and user experience changes in the complete AHC development cycle. This may be because of its complex nature, or maybe we wanted to give the best user experience for setting this up.

We thought that following may be the need of the user:

So, there were so many combinations users wish to evaluate. While developing the AHC product, we always wanted to give true freedom to our users. Users should be empowered enough to utilize the power of “simultaneous design evaluations”. We wanted to develop a user interface and workflows so that multiple design evaluations become a natural action from the user.

Considering all the above needs (and a lot of sleepless nights for our user experience team), we came up with a unique way of defining the design configurations and scenarios. The best part of this (and my personal favorite) is the Sankey diagram.

Configurations, scenario, and simulation

Sankey diagram is where you will see all your hard work defining the system. This will show all the options you wish to evaluate. And it does not stop there. These are pretty little boxes that know who their parents are and who their children are. You can select any parent and kids who come along the way. Although there are so many combinations to choose from, this unique representation makes it easy and fun to choose from.

Simulations and cloud

Ease of use and great user experience are two main aspects when we design any product. The third and probably the most important aspect for us at simulationHub is accessibility. We believe that “ideas can come anytime and anywhere”. To support this human nature, we need to have the required tools handy – anytime and anywhere. Imagine that I want to say something to my loved one, and it asks me to write a letter, post that and after a few days, someone will be able to read and feel what I felt like a few days back. A very simple and probably obvious example, but we have seen this exactly happening in the design environment. Imagine that I have a good design idea, I wait for the design model to be created, the simulation to be done, and maybe after a few hours (if you are lucky), you get what is the performance of your design. It's not different that your loved one receives the message after a few days.

At the inception of simulationHub itself, we decided on “anytime, anywhere”. This led to two aspects (which you will see in AHC as well as any other product developed on the simulationHub platform).

With this philosophy, we have the AHC product as a “cloud-based web product”, accessible to the user anywhere, anytime. So, once you select the simulations to run and submit, the algorithm starts a fleet of required computers, does the simulations, and comes back with the results. Meanwhile, it does so autonomously, you are free to do anything you want (or work on your next design idea).

Results

Once CFD calculations are completed, there is a huge amount of data generated. Velocity, temperature, and humidity at various locations in the domain are the basic parameters available. While conducting a CFD simulation, the complete region of interest is divided into millions of small volumes. After completion of the simulation, we will have each parameter at the center of such volumes, which means millions of numbers for a small region. Extracting the meaningful information from these numbers and presenting that in the form that we will be able to understand the performance is a challenge.

This is not a new challenge, and a lot of work has been done in the past. These CFD results are presented in the form of color plots, path lines, 3D clouds, etc. AHC product has these default result visualization methods available. This standard results representation may not help our users to understand the design performance quickly. So, the main challenge posed in front of us was to present the results in such a way that HVAC designers would understand that quickly and easily. Based on these results, the designer should quickly and accurately see what is working and what is not.

We have invented many unique ways for presenting the HVAC occupant thermal comfort data. It has all HVAC-specific parameters like PMV, PPD, EDT, DR available for visualization. We also have something called “comfort clouds” which gives quick and insightful information about comfort conditions in space.

My favorite in this category is the thermal comfort presented in the form of emojis. We thought of putting life (literally) in these results. We thought, what exactly a person at the place will feel when he is exposed to cold/hot air more than what is required for him. And if you see, it may talk to you by changing the expression. And technically, this is the real and quick truth about how HVAC system design is performing for occupant thermal comfort.

Thermal comfort emojis

Case studies

Nothing demonstrates the power of a product more than its real-world examples. View case studies of how AHC is used for estimating thermal comfort.

Closure (or opening for something bigger)

So, these are a few of the “Why’s and what’s” of AHC and its features. There are many such small stories behind each component available in AHC and I am sure we would love to talk about them. But let’s save them for later discussions. Now as we are launching the product, we wish to experience these stories. Please sign up for the app and give it a try. I am very sure; you will have something to say to us after your experience with the product.

Message

Subscribe Newsletter

App Dashboard

Autonomous Valve CFD

Autonomous Valve CFD

Autonomous HVAC CFD

Autonomous HVAC CFD