Development of an engine along with optimum Engine Control System is a fundamental and key stage of any vehicle pre-production program. There are three key stakeholders that come into play for vehicle engine development - Carmakers (OEMs), Customers and Industry & Government bodies. There is a conflict of objective with each of the stakeholders - OEMs want to cut cost, Customers want better Performance, and Industry and Government bodies want automakers to adhere to emission norms.
In addition, the constant time pressure makes powertrain development even more demanding. So is calibration method an ideal solution for the aforesaid challenges? Here are the excerpts from a conversation with Dr. Lakshman Hazra, Subject Matter Expert, Powertrain Practice, KPIT Technologies, on calibration methods for powertrain development.
What is the industry approach or the traditional methods used for
achieving engine calibration in test bench?

OEMs traditionally followed linear approach towards engine calibration measuring one parameter at a time. This led to increase in engine bench testing time.

In an engine test bench, one of the main tasks is to prepare maps for Engine Control Unit or Engine Management System (EMS). Traditional approach for map preparation involves capturing different engine input & output parameters on engine dynamometer test bench coupled with different measuring subsystems. A typical engine test bench consists of various measuring subsystems (engine oil, fuel supply, coolant, etc.) to operate the engine. Usually more than two engineers operate the test bench. A typical input and output engine measurement procedure consists of three simultaneous steps:
  1. Running or driving engine in different operating points by engine test bench computer
  2. Variation of Engine Control Unit’s parameter from ECU computer
  3. Measurement of engine input & output parameters in any of this
Sometimes maps are calculated from design data of engine and its subsystem, but verification in engine test bench is always done using engine dyno test bench.
What are the challenges in engine test
bench based calibration?

Increase in number of engine control parameters and measurement of those parameters is a big challenge for OEMs

As OEMs strive to meet the strict emission regulations, data analysis and map preparation are getting cumbersome with a large number of input parameters of gasoline and diesel engine. The main challenge in the traditional approach of measurement is the increase in number of engine control parameters such as engine power, fuel, smoke, emission, temperature – inside and outside the engine, oil cooling temperature etc. Making optimum map preparation using traditional method with these many parameters is very challenging. In addition, making interface for different complex measuring equipment and engine operational subsystem is a difficult task. High number of measurements indicate high cost for collecting data for map preparation.
How can model based calibration methods
help in optimizing engine parameters?

Design of Experiments (DoE) technique enables optimization of calibration process reducing the engine test bench time drastically.

Model Based Calibration (MBC) based Design of Experiment (DoE) technique is proven and commonly used solution for optimum map preparation of the engine. Here an optimum test plan is calculated on the basis of different optimum DoE algorithm (e.g. d-optimal, v-optimal, etc.) and measured only by this optimum test plan. The MBC Methodology has been depicted below:
Modeling used in MBC is statistical modeling and it is well known that statistical modeling takes a black box approach to capture the multivariable influence of inputs on outputs. Statistical engine models described here are steady-state models of the engine, because they account for a large fraction of the base calibration process. One more advantage of statistical modeling is that optimized map data can be generated depending on different criteria. User can add more test points by using any DoE method as validation points, which will be used for calculating statistical validation parameters of the calculated models. User friendly map editors are used to provide a better interface. KPIT has developed user friendly map editor and integrated it into this methodology as per the present industry requirements.
KPIT is known for providing leading technology solutions in the engine calibration space.
Can you tell us how KPIT helps various automakers in reducing the time required for powertrain development?
With over two decades of automotive electronics experience, we have a rich experience of working with the best names in the industry. We cherish our 12+ years of association with leading global Engine manufacturer and create value to provide reliable and high-performance solutions. Our journey as a strategic powertrain development partner with a leading OEM, demonstrated our technology leadership and helped OEM stay ahead of the competition. KPIT has developed Model Based Calibration Methodology (MBCM) and helped a renowned OEM by reducing their engine test bench time by 60% by using optimum map preparation. KPIT offers customized solutions such as development of in-house MBC tool as per client's OEM base map preparation. In a nut shell, KPIT can offer end to end consultancy for developing and implementing Model Based Calibration Methodology to engine makers & OEMs.

Terms of Use   Privacy Policy   Cookie Policy   Home   
Copyright © 2018 KPIT® Technologies Ltd.

CIN: L72200PN1990PLC059594