Simcenter Testing Solutions Simcenter Testlab Combustion Analysis

Combustion analysis is used to determine the performance and efficiency of a combustion engine.  By incorporating combustion analysis into Simcenter Testlab, the intent is to make understanding trade-offs between the noise and vibration produced by an engine versus the engine efficiency easier to understand.

The combustion analysis add-in in Simcenter Testlab is based on angle domain analysis of the cylinder pressure and volume.  In addition to multiple combustion metrics, the add-in includes a top dead center (TDC) determination tool and pegging capabilities.

The combustion analysis process is outlined in Figure 2 below:

There are several steps involved in doing combustion analysis which are covered in this article:
1. Combustion Analysis Background: PV Diagram, IMEP, PMEP, NMEP
2. Getting Started
   2.1 Required Signals
   2.2 Simcenter SCADAS
      2.2.1 Pressure Signals
      2.2.2 Encoder
      2.2.3 Torque
      2.2.4 Noise and Vibration
    2.3 Starting Simcenter Testlab
3. Angle Domain Validation
   3.1 Cylinder Alignment
   3.2 Angle Domain Conversion
   3.3 Zero or Top Dead Center (TDC) Angle Definition
      3.3.1 Manual Zero Definition
      3.3.2 Automatic Zero Definition
4. Combustion Analysis: Cylinder Geometry
5. Combustion Analysis: Top Dead Center (TDC) Determination
   5.1 TDC Accuracy Effect on Metric
   5.2 TDC Methods
6. Combustion Analysis: Pegging
   6.1 Reference Value
   6.2 Reference Channel: Intake Pressure
   6.3 Polytropic
7. Combustion Processing
   7.1 List of Combustion Metrics
       7.1.1 Indicated Mean Effective Pressure (IMEP)
       7.1.2 Pumping Mean Effective Pressure (PMEP)
       7.1.3 Net Mean Effective Pressure (NMEP)
      7.1.4 Brake Mean Effective Pressure (BMEP)
      7.1.5 Friction Mean Effective Pressure (FMEP)
      7.1.6 Mechanical Efficiency
      7.1.7 Peak Pressure
      7.1.8 Angle of Peak Pressure
      7.1.9 Peak Pressure Rise Rate (PPRR)
      7.1.10  Angle of Peak Pressure Rise Rate (PPRR)
       7.1.11 Burn Rates
       7.1.12 Combustion Noise (Lucas Combustion Noise)
   7.2 Data Storage
   7.3 PV Diagram

1. Combustion Analysis Background: PV Diagram, IMEP, PMEP, NMEP

What is combustion analysis exactly?  It is an analysis that is performed over a complete engine cycle (intake-compression-power-exhaust).  Using the pressures generated in a cylinder of engine and the volume within the cylinder, the amount net energy created during a complete cycle can be calculated. 

The Net Mean Effective Pressure (NMEP) is one of the key metrics calculated from combustion analysis.  It is an indication of the work output for the engine.

Net Mean Effective Pressure (NMEP) and related combustion metrics are calculated by plotting the pressure during a combustion cycle against the volume within the cylinder as shown in Figure 3.
 

Using the Pressure-Volume (PV) diagram, two areas are formed as shown in Figure 4:
 

Figure 4: Pressure (Y-axis) versus volume (X-axis) from one complete combustion cycle.  The volume is plotted from Top-Dead-Center (TDC) position of the piston in the cylinder to the Bottom-Dead-Center (BDC) position. The IMEP (green area) is proportional to the work done while the PMEP (red area) is proportional to losses.  The net difference between IMEP and PMEP indicates the efficiency.
 

Three main metrics that can be calculated from the areas on the PV diagram:

There are several other metrics that can be calculated in Simcenter Testlab Combustion Analysis (including combustion noise via structural attenuation, peak pressure rise rate, mechanical efficiency, etc).

2.  Getting Started

The required measurements, hardware, and software to perform combustion analysis in Simcenter Testlab is covered in this section:

2.1 Required Signals

The following are measured to perform combustion analysis:

Additionally, the dimensions of the cylinder/piston must be known to calculate the volume within the cylinder as the piston moves up and down. The firing order of the cylinders in the engine must also be known for cylinder alignment, for example for a 4-stroke 4 cylinder engine, the commonly used firing order is 1-3-4-2.

2.2 SCADAS Hardware

Simcenter SCADAS hardware has some specialized inputs to help measure the encoder and pressure signals:

2.2.1 Pressure Signals

Typically, combustion pressure measurements use piezo electric pressure sensors which are flush-mounted in the cylinders.  These sensors deliver charge output.

They can be recorded with a Simcenter SCADAS in one of two ways:

2.2.2 Encoder

To perform combustion analysis, two encoder signals are needed:

Most SCADAS units are equipped with two tachometer inputs that can convert up to 40,000 pulses per second into rpm. These two inputs can be used to measure the reference and multiple pulse signals.

A device called an incremental encoder is an alternative to measuring two encoder signals separately.  An incremental encoder contains both a reference pulse and two multiple pulse per revolution signals in one device (Figure 5):
 

The Simcenter SCADAS RV4 card can accommodate an incremental encoder.  The two multiple pulse per revolution signals can be decoded by the RV4 card to determine the direction of rotation. Although the sense of rotation is not really needed for combustion analysis, it is helpful for torsional vibration measurements.

2.2.3 Torque

If it is desired to calculate the Brake Mean Effective Pressure (BMEP) combustion metric, a strain-based torque sensor in a flex plate can be used. The flex plate is installed on the fly wheel and uses telemetry to transmit the torque signal.  The receiver needs to convert the signal to an analog voltage to be read into the Simcenter SCADAS hardware.

2.2.4 Noise or Vibration

Any noise or vibration data of interest is also recorded in parallel, typically using ICP/IEPE based accelerometer or microphones  

This article assumes the measurements are already acquired to perform the combustion analysis.  It is possible to also run the tool during live acquisition, but this is not covered in the article.

For more information on acquiring data with Simcenter Testlab, see the following articles:

2.3 Starting Simcenter Testlab

To run combustion analysis, go to “Tools -> Add-ins” and turn on Combustion Analysis as shown in Figure 6.
 

If using Simcenter Testlab token licensing, there are 98 tokens total required: Combustion Analysis (20 tokens) requires Angle Domain Analysis (42 tokens), Signature Throughput Processing (36 tokens), and the Customized Metrics Calculator (zero tokens). 

More information on Simcenter Testlab token licensing in this knowledge article: Simcenter Testlab Tokens: What are they, and how do they work?

3. Angle Domain Validation

Before performing the combustion analysis, the measurement data needs to be converted from a time basis to angle basis.  This is done in the Angle Domain Validation worksheet (Figure 7).
 

In the angle domain worksheet, the data is prepped: combustion events are aligned between cylinders, the revolutions per cycle are set, the angular resolution is fixed, and the zero-degree mark (denoting Top Dead Center position of the piston in the cylinder) is determined.  These steps are highlighted in Figure 8 below:
 

Figure 8: Combustion analysis process starts with the Angle Domain Validation workbook (green).
 

These steps are explained in detail in the next sections.

3.1 Cylinder Alignment

If the engine has multiple cylinders, alignment of the pressure pulsations needs to be performed between the different cylinders.

A “Cylinder Offset” is entered for each pressure trace in the upper left of the Angle Domain Validation worksheet as shown in Figure 9:
 

Figure 9: Combustion occurs in each cylinder at a different angular position (upper left), using the “Cylinder Offset” field in the Angle Domain Worksheet (middle), the combustion events are aligned between cylinders (lower right).

 

Normally the cylinder alignment is set to zero degrees (scroll to right to be able to see the “Cylinder Offset” field).  After entering the cylinder alignment correctly, when displaying the angle domain data the pressure peaks between cylinders should occur at the same position in all traces when overlaid. 

This is done to needed to perform the combustion calculations.

Note that "Cylinder Offset" is a field that can be specified in Channel Setup during acquisition as well.

3.2 Angle Domain Conversion

Pressures, vibration, and sound measurements by the Simcenter SCADAS hardware are acquired as a function of time. Using the high-resolution encoder information, the time-based data is transformed to angle domain data.

After the time recording (with pressures, tachometer, etc) is selected in the Time Data Selection worksheet, the Angle Domain Validation worksheet is used to convert the time data to angle domain data. This step makes the data that is equidistant in time into equidistant in angle. It is done by adaptive resampling of the data.

In the lower left corner of the worksheet, there is a “Angle Domain Definition” area (Figure 10):
 

Figure 10: Angle Domain Definition menu.
 

The following can be set:

3.3 Zero or Top Dead Center (TDC) Angle Definition

To perform combustion analysis, the angular position where the piston is at its farthest position of travel in the cylinder needs to be marked as zero degrees.  This is called the “Top Dead Center” or TDC position of the piston within the cylinder as shown in Figure 11.

For a non-combusting engine, the TDC position typically corresponds to where the pressure within the cylinder is at its peak.  This could vary if a thermodynamic loss angle (usually less than one degree) adjustment is used. When an engine is combusting, the peak pressure can occur 10 to 20 degrees after the TDC position.

Using a single cycle, the zero position can be set for TDC in Angle Domain Validation. There are two areas (ie, two methods) that can be used to set zero-degree position of the combustion cycle as shown in Figure 12:
 

Figure 12: In the lower left of Angle Domain Validation worksheet, there are two different methods for setting the zero-degree position: Automatic (left side menu) or Manual (right side menu).
 

After setting the zero for a single cycle, there are further statistical checks on the TDC position done on all the cycles in the Combustion Analysis add-in.  This is explained in an upcoming section.

The methods within Angle Domain Validation are for setting the zero angle are further explained here:

3.3.1 Manual Zero Definition

If the TDC position is known by design, the angle offset can be directly entered manually in the Angle Domain Validation sheet. In this case, it is not necessary to use the TDC options in the Combustion Analysis (covered in upcoming section).

Option 1 – Direct Entry: If the data was acquired using an incremental encoder and a Simcenter SCADAS, a reference pulse (one pulse per revolution signal) is available.  The reference pulse location can be shown by changing “rpm” under the “RPMPulse” field to “Ref Pulse” as shown in Figure 13.
 

By default, the first reference pulse is set to zero degrees.  This many not correspond to TDC.

The angle offset between the reference pulse to the TDC angle position can be directly entered in the “Zero Angle Definition” as shown in Figure 14.

In the example above, the zero aligns with a reference pulse trace after pressing “Refresh Data in Display”.

Option 2 – Manual Cursor: If a reference pulse is not available, a cursor can be used to set the zero angle to correspond to TDC.  Go to “Manual Zero Angle Definition” menu tab and select “Show cursor” (Figure 15).  
 

Position the cursor to the 0:0 position (where TDC is expected), enter the “Expected Angle Value” and press “Apply”.

The displays will update with the zero-angle reset to the selected cursor position (Figure 16).
 

Note: Using this method the TDC offset can be defined based on any input channel, not necessarily the reference pulse.  

3.3.2 Automatic Zero Definition

Automatic zero definition can be useful for measurements performed on 4-stroke engines that require two rotations of the crankshaft for a full cycle. 

For example, the reference pulse occurs once per revolution. The reference pulse position only corresponds to a power stroke (cylinder pressure spikes – blue line) of the engine every other revolution (Figure 17).
 

When processing / visualizing in angle domain it is very important to define which rotation corresponds to the beginning of the cycle.  This is difficult in the four-stroke engine example unless another trace, like the pressure trace (blue line) is also considered.  The pressure spikes high on the power stroke and has considerably less pressure build up on non-power stroke.

To use the pressure trace to determine which pulse is the correct zero or TDC reference, press the “Advanced…” button as shown in Figure 18.
 

More information on the Angle Domain Validation worksheet in the knowledge article: Angle Domain: What is it?.

4. Combustion Analysis: Cylinder Geometry

The next step in the combustion analysis is to enter the cylinder geometry information (Figure 19):

Figure 19: Cylinder geometry step (highlighted in blue) is the first step in the Combustion Analysis workbook (yellow).
 

The cylinder geometry information is used to calculate the volume of air in the cylinder.  The volume is calculated as a function of the engine cycle.  This is done by pressing the “Cylinder Geometry…” button in Simcenter Testlab Combustion Analysis (Figure 20) workbook.
 

It is required to fill in the engine parameters accurately to calculate the P/V diagram and all related metrics using volume information.

The swept volume is defined as follows in Figure 21:
 

Figure 21: Swept volume formulas and definition for cylinder volume.
 

Entering the information results in a trace of the volume versus engine cycle (Figure 22) that is used in all calculations. 
 

Some further notes:

5. Combustion Analysis: Top Dead Center (TDC) Determination

Further validation of the TDC/zero position is done in the combustion analysis add-in.  In the “TDC Zero Angle…” area of Combustion Analysis, the zero-angle determination will be checked and corrected as needed.  This step is highlighted in Figure 23 below:

Using the settings defined in the Angle Validation for one cycle, the Top Dead Center calculator (Figure 24) in Combustion Analysis worksheet will perform a cycle-by-cycle inspection on all the cycles of the pressure data. 
 

The Top-Dead-Center (TDC) analysis is usually performed on a motoring engine as shown in Figure 25.
 

A motoring engine (no fuel, starter or other used to turn the engine) is used because the measured peak pressure in the cylinder corresponds to the TDC position of the piston.  In a running engine, the peak pressure TDC is usually after the peak pressure due to thermodynamic effects (Thermodynamic Loss Angle: ~0.7 deg) of the combustion event.

When cylinder pressure in a motored engine is used to calibrate the TDC, it is commonly advised to use the cylinder immediately next to the crankshaft encoder to limit the torsional vibration influence. 

5.1 TDC Accuracy Effect on Metric

A zero angle or TDC should ideally be determined within 0.1 degrees.  The accuracy impacts the estimation of the cylinder volume and many derived metrics like IMEP.
The effect of a +/- 10 degree offset on the TDC is shown below for one combustion cycle in a PV diagram in Figure 26 below:

Figure 26: Effect of TDC accuracy on PV Diagram. Left: PV diagram with +10 offset in TDC (blue), correct TDC (red), -10 offset (green).  Right: Combustion metrics IMEP/PMEP results 10-15% different due to 10 degree difference.
 

The effect of a 10 degree error can change the IMEP, PMEP, and NMEP and other combustion metrics by 10 to 15%.

If only predicting combustion noise with no interest in other combustion metrics, the TDC determination is less critical.  As long as the complete pressure pulsation is contained in the angle trace, the combustion noise prediction will not be affected.

5.2 TDC Methods

There are two methods in the TDC menu for checking the accuracy:

The TDC menu is the worked from top to bottom (Figure 27):
 

 

Some of the major steps in the TDC menu include:

The Top Dead Center calculator calculates an average Top Dead Center position from all the cycles as well as the standard deviation of the positions.  Any problems or inconsistencies in the data is uncovered in this manner.

If the calculated TDC is more accurate, pressing the “Use for Processing” button updates the value being used.  This is the same as setting the zero angle value in Angle Domain Validation.

6. Combustion Analysis: Pegging

When measuring combustion pressure there is often drift in the pressure signal. High temperatures inside the cylinder induce drift in the pressure transducer.
The drift is removed in the “Pegging” step of Combustion Analysis (Figure 28):
 

This drift is removed from the pressure signal to calculate the combustion efficiency accurately.  The desired outcome of the “pegging” step is illustrated in Figure 29.
 

Pegging is performed separately on each combustion cycle. Removing the drift ensures that the combustion is based on the change in pressure over the cycle and not distorted by any offset/drift in the data. 

Pegging can be accomplished in one of three ways in Simcenter Testlab Combustion Analysis (Figure 30):
 

The pegging options are described next:

6.1 Reference Value

If the pressure offset is constant and does not vary over time in the recording, then one single fixed value can be used to eliminate the offset. This is typically valid for constant speed and load conditions.

6.2 Reference Channel: Intake Pressure

The reference channel method requires an additional pressure measurement: an additional intake manifold pressure on top of the individual cylinder pressure measurements.  This method adjusts the pressure offset separately on each combustion cycle. This accounts for varying offsets withing a single recording.

At some point during the combustion cycle, the pressure in the intake manifold and the pressure in the cylinder are equal. Comparing the pressure between those two locations at the right cycle moments allows compensating for the drift on the cylinder pressure as shown in Figure 31 
 

The determination of the best cycle moments (few degrees before or after Inlet Bottom Dead Center – IBDC) for the whole load/speed range of the engine is based on experience.

6.3 Polytropic

The compression stroke of a four-stroke engine, after the intake valve has closed and before the combustion has started, is accurately represented by a polytropic compression curve. This property can be used for pegging. The polytropic compression can be defined by Equation 1:

Equation 1: Polytropic Constant

 

Where:

If the sensor measurement is denoted pm, and the offset is denoted p, the cylinder pressure (p) can be expressed per Equation 2 below:
 

Combining the previous two equations yields Equation 3:
 

Equation 3: Polytropic pegging equation.
 

Assuming K is known, there are two unknowns, Δp and C. These unknowns can be precisely determined using two matching pairs of pressure/volume measurements. With more pressure measurements, the equations can be solved in a least squares sense.

Note that for air, a polytropic constant of 1.33 is typical.

7. Combustion Processing

Combustion metrics (Figure 32) can be calculated after the TDC position is determined, pegging was performed, and cylinder geometry has been entered. 
 

Metrics to be calculated are selected on the right-hand side of the combustion add-in as shown in Figure 33.
 

Figure 33: Available combustion metrics.in the Simcenter Testlab Combustion Analysis workbook.
 

Select the metrics to calculate. Press “Apply and Calculate” button when finished selecting the metrics.

Some of the available metrics are explained here:  

7.1 List of Combustion Metrics

The following metrics can be calculated in Simcenter Testlab Combustion Analysis:

7.1.1 Indicated Mean Effective Pressure (IMEP)

IMEP is a measure of the thermodynamic work done by the gas on the piston face during the engine cycle divided by the swept volume. It can be seen as the average pressure during the engine cycle.

7.1.2 Pumping Mean Effective Pressure (PMEP)

PMEP corresponds to the losses due to the pumping effect during exhaust and intake phases.

7.1.3 Net Mean Effective Pressure (NMEP)

NMEP corresponds to the resulting effort from the piston on the crankshaft. It is the net difference between the pumping effective mean pressure and indicated mean effective pressure.

7.1.4 Brake Mean Effective Pressure (BMEP)

Work delivered by the crankshaft is less than the indicated work due to mechanical friction and parasitic loads of the engine. Parasitic loads include the oil pump, supercharger, air conditioner compressor, alternator, etc. Actual work available on the crankshaft is called brake work. The brake estimated from the torque (τ) measurement on the output shaft of the engine

To be calculated, the metric requires a DC torque measured at the output of the engine. Formulation is shown in Figure 34:
 

FDR is the Final Drive Ratio and the τ(i) measured torque. The FDR depends on the location of τ(i) measurement.

7.1.5 Friction Mean Effective Pressure (FMEP)

The Friction Mean Effective Pressure (FMEP) is the difference between Net Mean Effective Pressure (NMEP) and Brake Mean Effective Pressure (BMEP):
 

A single value per cycle is calculated.
 

7.1.6 Mechanical Efficiency

The mechanical efficiency is the ratio of the brake work at the crankshaft to the indicated work in the combustion chamber:
 

A single value per cycle is calculated.
 

7.1.7 Peak Pressure

Maximum pressure in cylinder over one engine cycle.  Single value per cycle.

7.1.8 Angle of Peak Pressure

Angle at which maximum pressure occurs in cylinder over one engine cycle. Single value per cycle.
 
7.1.9 Peak Pressure Rise Rate (PPRR)

The peak pressure rise rate is the maximum increase of the pressure during the cycle expressed in unit of pressure over degree (Example: bar/deg).

 

Higher rise rates usually have adverse effect on noise and vibration.  

Single value per cycle.

7.1.10  Angle of Peak Pressure Rise Rate (PPRR)

The angle of peak pressure rise rate is the angle at which the pressure increases the fastest during the cycle.
 

 

Single value per cycle.

7.1.11 Burn Rates

The mass fraction burned is the percent of total fuel burned at given crank angle. It is defined as:
 

 

Where:

The burn rates correspond to the angle at which a defined % of the fuel has been burned (Figure 35):
 

7.1.12 Combustion Noise (Lucas Combustion Noise)

The combustion noise is the noise radiated by the surface of the engine structure as it reacts to the rapid pressure increase in the cylinder(s). However, it is not possible to derive the combustion noise directly from the pressure rise. Different calculations have been implemented, like the Lucas filter methodology (Figure 36):
 

Experimentation has proven that the attenuation properties of most engine structures are very similar, and a standard average attenuation curve was proposed (SA1 curve) but the recent engine design change makes that this standard deviates from current situation. The structural attenuation window is therefore user defined: any frequency block can be selected.

An example attenuation curve is shown in Figure 37.
 

In Simcenter Testlab, the combustion curve is unitless (unit is “/”) and in 1/3 octave format.  The cylinder pressure is divided the attenuation curve.  In the figure above, notice that the curve is somewhat the inverse of the  A-weighting curve of human hearing.  Low and high frequencies will be reduced.

The result from the Lucas combustion noise metric is the calculated noise produced by the combustion in a specific cylinder at one meter from the engine.

7.2 Data Storage

The combustion metric data is stored in several folders within a new run in the Simcenter Testlab project.

One of the folders is called “User”, which is short for “User section” as shown in Figure 38:
 

The user metric data contains the results of any metric calculation that creates one value per engine cycle.  For example, the PMEP, NMEP, and IMEP values for 1000 cycles are plotted in Figure 39:
 

 

The complete angle-based data for each engine cycle is also stored in the run.  These are stored in the Waterfall area in the Derived folder as shown in Figure 40:
 

Figure 41 is a plot stored in Waterfall/Derived folder for one engine cycle.
 

The pressure and volume can be plotted against each other as explained in the next section.

7.3 PV Diagram

To plot the pressure and volume of an engine cycle against each other, open a “CA_XY” graph as shown in Figure 42:
 

A PV diagram can then be made by plotting the Volume channel on X-axis and the Pressure data on Y-axis as shown in Figure 43:
 

Click into the waterfall data to select a single cycle to display in the graph.  There is waterfalls of both pressure and volume, with the cycle annotated (5:0 mean 5th cycle starting a zero degrees).

More information in the knowledge article: Simcenter Testlab Displays: XY Graph


Hope this helps with combustion analysis and combustion noise prediction!

Questions?  Email peter.schaldenbrand@siemens.com

Related Links:

• Index of Simcenter Testing Knowledge Articles
• What is an order?
• Torsional Vibration: What is it?
• Tips and Tricks for Acquiring Torsional Orders
• Zebra Tape Butt Joint Correction for Torsional Vibrations
• Measuring RPM: Missing Pulses
• Balancing: Static, Coupled, and Dynamic
• Removing Spikes from RPM Signals
• Using the Smoothing Function to Remove RPM Spikes
• RPM Calculation Problems and Laser Tachometers
• Simcenter Testlab Signature
• Fixed Sampling versus Synchronous Sampling
• Angle Domain Analysis
• Harmonic Removal
• Simcenter Testlab: Colormap Displays
• Interpreting Colormaps
• Orbit Plots
• Cycle to Cycle Averaging in Simcenter Testlab
• Speed Sweep Data Processing: RPM Increment, Framesize, Sweep Rate, and Overlap
• Order Cuts: How to Get the Correct Amplitude?
• Knock Detection Algorithm Development
• Simcenter Testlab: Recalculating Levels with Orders Removed
• Human Body Vibration
• Simcenter Testlab: Switching Frequencies and PWM Signals
• Simcenter Testlab: Calculating Angular Difference
• Rotating Machinery Dynamics Seminar
• Gears Rotating Machinery Dynamics Seminar
• Rotating Machinery Testing YouTube Playlist