|INSIGHT: Improving Design Robustness and Agility with Cloud-Based CFD
Cloud computing has the potential to redefine the way CFD is used and licensed for turbomachinery aero design. In this issue of The Flow, we sit down with Bob Ni to explore cloud computing and how companies can capitalize on this trend for improved design robustness and agility. Bob is the Chairman and CTO of ADS. Prior to founding the company, he was a senior fellow at Pratt & Whitney leading turbomachinery CFD development and application efforts.
FLOW: What is cloud computing?
BOB: I like to think of cloud computing as a web service that gives organizations on-demand access to computing infrastructure on a pay-as-you-go basis.
FLOW: How can cloud computing help with turbomachinery aero design?
BOB: When paired with CFD, cloud computing gives aero designers instant access to massive CFD analysis capacity, all of which can be scaled up/down as needed and paid for on a metered basis.
This combination of capacity on demand and pay as you go pricing is extremely potent for aerodynamic design. It allows organizations to use scale to compress analysis cycles for greater design robustness and agility, without the need for large upfront investment in hardware and CFD licenses.
FLOW: Aren't you still gated by CFD licensing constraints?
It depends on your situation. This may very well be the case if you are using commercial codes with traditional lease or perpetual licensing; check with your vendor to understand your options. On the flip side, it should be no issue if you are using non-commercial codes or taking advantage of commercial CFD offerings designed for the cloud like Cloud Leo
, where even the CFD licensing is priced on a pay-as-you-go basis.
FLOW: Interesting. Can large and small businesses take advantage of cloud-based CFD?
WILL: Yes, cloud-based CFD benefits organizations both large and small, albeit in different ways. For large organizations, cloud-based CFD can supplement their large internal clusters with short term “surge” capacity to handle peak loads during critical phases of the design cycle.
For small businesses, the impact is more pronounced. Most of these organizations lack the resources to invest in and maintain large clusters. As a result it is often their cluster size or CFD license that limits their ability to thoroughly analyze a design. Cloud-based CFD removes this constraint, allowing small businesses to provision as much analysis capacity as needed to do aero design right, and pay only for capacity consumed. The scale, flexibility and economics of cloud-based CFD helps small businesses improve design robustness and agility. And level the playing field against larger competitors.
FLOW: What are some examples of the possibilities with cloud-based CFD?
BOB: Let’s say you’re an independent consultant—you’ve spent 15-20 years at a Tier 1 OEM and have recently struck out on your own, providing compressor aero design and analysis services to your clients. From your time working at the OEM, you know what’s possible with CFD, but you certainly do not have access to anything near the analysis resources you had from those days.
So what do you do? You compromise. You acquire an eight-core workstation, bite the bullet on a $25,000-$35,000 CFD license and come up with simplifying approximations to make your analysis work within the time and budget constraints of your project proposals.
Cloud-based CFD flips this notion on its head. With cloud-based CFD, analysis capacity is provisioned as needed and paid for on an as-you-go basis. No large upfront investment in hardware and CFD licenses is required, just an internet connection and a credit card to pay for the CFD analysis resources consumed. Better yet, there are no artificial capacity limits constraining the number and size of cases you can run; you simply provision as much analysis resource as needed and pay only for what you consume.
FLOW: So it sounds like cloud-enabled CFD helps to reduce the upfront cost of aero analysis?
BOB: That's an important benefit but not the most important. The most important is that aero designers can now to apply massive scale to compress analysis cycles for better design robustness and agility. For example, what if you could produce a full compressor map in 2.5 hours for under $600? Or carry out full wheel unsteady analysis for durability design over a weekend for less than $2,000? The possibilities are endless and limited only by your expertise, not by your CFD license. Check out the case study in the next article below to see an example of cloud-based CFD at work.
FLOW: Is cloud-based CFD secure?
BOB: If you use vendors like Amazon Web Services, cloud -based CFD is likely to be as secure as most large company internal networks, and far more secure than running it in small business settings. Amazon has many years of experience designing large data centers and has comprehensive policies in place to ensure physical security, secure access to services and data privacy.
FLOW: Sounds interesting, how can our readers get started?
Check with your current CFD vendor to see if cloud computing is supported. You can also find out more about Cloud Leo at www.cloudleo.com
. We have a beta program in place through June 30, 2013 that will give participants an easy way to experience the possibilities firsthand for 60 days for nothing except the cost of an Amazon Web Services account.
FLOW: Thanks, Bob.
BOB: My pleasure.
CASE STUDY: Fast and Cost-Effective Compressor Map Generation with Cloud-Based CFD Cloud-based CFD enables compressor aero designers to generate compressor maps quickly and cost-effectively, without large investment in cluster hardware and CFD licenses. In this case study, the cloud CFD service Cloud Leo is used to demonstrate how a 39 point compressor map for the NASA CC3 centrifugal compressor can be generated in 2.5 hours and for less than $600. <more>
TECHTIPS: Generating a Fast Fourier Transform on Unsteady Surface Loading Data for Forced Response Prediction The latest release of ADS CFD (5.6) now includes the ability to generate a Fast Fourier Transform (FFT) on time accurate surface loading data to aid in forced response prediction. In this tech tip, we'll show you how to use the utility ADS-UTIL to generate the FFT, plot the power spectrum and show the Fourier coefficient contours on the airfoil surface using ParaView. <more>
Welcome to The Flow
Welcome to The Flow, a newsletter for monthly insights on turbomachinery CFD published by AeroDynamic Solutions, Inc.
Each month we'll spotlight a topic of interest, discuss a case study and/or provide useful pointers about how to get the most out of the ADS CFD system.
You are receiving this email because you or someone else you know thought you would be interested. To unsubscribe please click here. We value your privacy.