Particulate mass measurements
of diesel exhaust gases
RAFAEL VILLASMIL
Particulate mass measurements
of diesel exhaust gases
Rafael Villasmil
Master of Science Thesis MMK 2006:16 MFM94 KTH Industrial Engineering and Management
Examensarbete MMK 2006:16 MFM94 Mätning av partikelmassa i avgaser
från en dieselmotor Rafael Villasmil Godkänt 2006-02-14 Examinator Hans-Erik Ångström Handledare Hans-Erik Ångström Uppdragsgivare KTH Kontaktperson Hans-Erik Ångström Sammanfattning
Detta examensarbete är en fortsättning på en tidigare studie inom karakterisering av dieselavgaser och utgör en del av det akademiska arbetet inkluderat i EMIR-1 projektet. Huvudmålet med examensarbetet är att slutföra installationen och utföra lämpliga modifieringar av ett mätsystem som finns i en av motortestcellerna på KTH för att kunna mäta partikelmassan i avgaserna från en dieselmotor.
Mätning av partikelmassa med instrumentet TEOM kräver att provet som tas från avgaserna är utspätt. Under första delen av detta projekt var målet att studera och utföra lämpliga modifieringar av den existerande KTH-spädaren eftersom tidigare tester visat att det fanns ett flertal problem att få ett konstant spädningsförhållande från spädaren.
Dessa modifieringar gjordes och resulterade i en tydlig förbättring av det maximala in- och utloppsflöde som spädaren klarar. Ytterligare modifikationer har även föreslagits för att kontrollera inloppsflödet så det blir möjligt att reglera spädningsförhållandet och göra det konstant över ett brett register av motorvarvtal och motorlast.
Den andra och viktigaste delen var att göra partikelmassmätning möjlig genom att analysera frekvenssignalen från TEOM-instrumentet. Idén var att få ut frekvensen från den oscillerande delen i instrumentet och att därav med de rätta korrelationerna kunna bestämma masskoncentrationen av partiklar i avgasflödet från motorn.
Ett elektroniskt problem i TEOMens krets upptäcktes och korrigerades, så att den frekvensbärande signalen från TEOMen kan analyseras på ett korrekt sätt. Frekvenssignalen kan efter modifikationen mätas under ett tidsintervall med försumbara störningar. På detta sätt har ett effektivt sätt att bestämma partikelmasskoncentrationen i avgaser tagits fram. För att förbättra metoden bör dock ett par problem åtgärdas. Kondensering av vatten och kolväten inuti rören påverkar mätningarna och bör undvikas genom värmning av provtagningssystemet. Idén är att få ett kraftigt utspätt och torrt flöde.
Som en fortsättning av detta arbete har ett nytt examensarbete initierats. Målet med det kommande arbetet är att implementera andra spädningsutrustningar såsom spädaren med roterande skiva och också en partikelräknare. Partikelräknaren ska fås att fungera med KTH-spädaren i ett uppvärmt provtagningssystem för att undvika kondenseringsproblem.
Master of Science Thesis MMK 2006:16 MFM94
Particulate mass measurements of diesel exhaust gases
Rafael Villasmil Approved 2006-02-14 Examiner Hans-Erik Ångström Supervisor Hans-Erik Ångström Commissioner KTH Contact person Hans-Erik Ångström Abstract
This master thesis was a continuation of a previous study in the diesel exhaust characterization, which is part of the academical work embedded in the EMIR-1 project. The main objective of this thesis work was to finish the installation and make the proper modifications of the particulate mass measuring system located in one of the engine test cells at KTH, in order to be able to measure the mass of the particles from the exhaust of diesel engines.
This experiment requires a diluted flow sample from the exhaust that should be sent to the device that makes possible the measurement of particle mass, called TEOM. In the first stage of this project, the objective was to study and make the proper modifications in the existing KTH diluter because there were several problems to obtain a constant dilution ratio from it during tests.
This modifications were made and it resulted in a great improvement in the maximum inlet and outlet flow that this diluter can handle. Therefore another modification is proposed for controlling the inlet exhaust flow so it will be possible to regulate the dilution ratio and make it constant along a wide range of engine loads and speeds.
The second and most important stage was to make particulate mass measurements possible by analyzing the frequency signal from the TEOM device. The idea was to obtain the frequency of oscillation from the tapered element in the TEOM, and therefore with the proper correlations, be able to predict the particulate mass concentration in the exhaust flow from the engine.
An electrical problem in the TEOM circuit was detected and corrected, so the signal that carries the frequency from the TEOM can be analyzed properly. The frequency signal can be measured in a time interval with minor disturbances. In that way, it has been found an efficient way to predict the particle mass concentration in the exhaust.
Although this method is efficient, there are still some issues yet to be solved. The condensation of water and hydrocarbons inside the pipelines that affects the particulate measurements should be avoided by heating the sample flow. The idea is to get the most diluted and dry flow.
As a continuation of the present master thesis, a new thesis work has been initiated. The purpose of this future work is to implement other dilution devices as the rotating disk diluter and also a particle counter, to make it work together with the KTH diluter operating with heated lines, to avoid condensation problems.
Contents
1 Introduction 2
2 Sampling system description 3
2.1 Engine test cell 3
2.2 Dilution devices 4
2.2.1 Dekati Ejector diluter 4
2.2.2 KTH Ejector diluter 5
2.2.2.1 KTH diluter modifications 6 2.2.3 Dump volume for avoiding pressure pulsations 10 2.2.4 Diluter with heater, system (Proposal) 11
2.3 Particulate mass measurement devices 12
2.3.1 Tapered Element Oscilating Microbalance (TEOM 1100) 12 2.3.1.1 Calibration of the TEOM device 15 2.3.1.2 Electrical modification in TEOM circuit 17
2.3.2 Smokemeter 20
3 Test Procedures 20
3.1 Flow dilution tests 20
3.2 Particulate mass concentration tests 22
4 Results 23
4.1 Flow dilution results 23
4.1.1 Dekati dilution ratio test 23
4.1.2 Dekati dilution ratio with dump volume test 24
4.2 Particulate mass measurements 25
6 Discussion 30
7 Conclusions 30
1. Introduction
The particulate matter (PM) emitted by combustion engines is a problem which affects people worldwide. It can be the cause of many diseases including cancer. Therefore, the government of Sweden has taken action to develop better engine designs based on the studies and researches made by both academical and industrial sides.
The particle formation can be affected by different factors, such as the fuel quality, the design of the combustion chamber and the environmental conditions. Tests have to be done under controlled experimental conditions to avoid the different phenomena that can occur with particles. An example of these problems is the condensation formed in the pipes of the sampling, originated by nucleation phenomena, where the gas species adhere to particles, and form the condensate that affects the sample. This nucleation can be avoided by using dilution devices. The high concentration of particles can be avoided by the use of particulate traps.
The research has to be continuous, and the new measuring techniques must be stable to sampling temperature, must show a constant dilution ratio, and should contemplate the residence time of the particles in order to get appropriate data from the tests. [1]
In this thesis work, the main objective was to develop a sampling system for measuring one of the most important parameters of these particles, their mass concentration in the exhaust flow from a Scania DC 1201 diesel engine. For that purpose, a diluter was built in the previous thesis work, to dilute the flow going to the mass concentration device, the tapered element oscillating microbalance, or TEOM.
This device has a tapered element that vibrates as the filter it has attached to it accumulates particles from the sample flow. By measuring the frequency of oscillation of the TEOM it can be predicted the mass concentration of the particles in the exhaust of the engine.
The built diluter, referred to as KTH diluter, has shown some problems to maintain a constant dilution ratio during tests. In this research will be presented the results from the modification of the KTH diluter, and also the changes made to the electronic circuit of the TEOM device, in order to be able to measure the particle mass from the exhaust of the diesel engine.
2. Sampling system description
The sampling system was initially developed in the previous work by Jesper Björkstrand, and consisted in installing the necessary components for taking a sample of the exhaust flow from the Scania DC1201 engine, dilute it with air and send it to the TEOM instrument for making the particulate mass measurements.
However, some of the elements installed have been modified in the present work. Also the configuration of the sampling system was changed, adding more elements to the system. This resulted in different configurations depending on the desired measurement.
2.1 Engine test cell
Figure 1 shows the final configuration of the sampling system for measuring the particulate mass.
Possibility to install after treatment device
Figure 1: Schematic of the particulate mass measuring system.
The system has two possibilities of taking the sample from, controlled by manual valves (MV). This allowed taking samples from upstream and downstream of the particulate filter when it was installed in the system. The particulate filter was removed from the system and for the measurements done in this research, the sample flow was collected by setting open MV-1 and MV-2, for leading the flow to the ejector diluter and returning the excess air mix to the exhaust pipe. MV-3 and MV-4 remained closed.
The pressure regulator (PR) set the pressure for the air connected to the diluter to 2 bar (relative pressure). The air is cleaned of particles with a HEPA filter that assures the air to be free of particles before mixing with the exhaust flow.
The engine cell’s computer controls the electromagnetic valves (EMV). EMV-1 allows turning on/off the pressurized air through the diluter. EMV-2 and EMV-3 are connected inverse, so when one is open, the other is closed. Because of this configuration, the TEOM test can be performed while the engine is running. When the TEOM is on but no sample is taken from the exhaust, the TEOM vacuum pump sucks clean ambient air that has passed through a HEPA filter.
2.2 Dilution devices
The samples that are taken from the exhaust flow of the engine cannot be conduced directly to the particulate mass measuring devices, because they are highly particulate concentrated. Also the sample might be affected because of the high temperatures and pressure pulsations that are present in the engine exhaust. [2]
Therefore, the sample from the engine flow must be diluted first in order to make further particulate mass measurements. Two devices for diluting the exhaust flow were studied during this research. The objective was to establish a comparison between both ejector diluters: the Dekati diluter and the KTH diluter. Since the KTH diluter had problems in previous tests with keeping a constant dilution ratio, both diluters were studied for identifying the probable causes of the dilution ratio variation. From this comparisons, some modifications were made to the KTH diluter with the purpose of improving the previous ejector diluter design.
2.2.1 Dekati ejector diluter
Figure 2: Sample flow through Dekati diluter. [3].
The operating principle of this diluter is based on under pressure caused by the dilution gas flowing through an annular nozzle. This under pressure draws a sample flow through this nozzle. If the dilution gas flow is increased the under pressure increases correspondingly and the sample flow is increased (Figure 3). Thus, the dilution ratio remains constant, if the pressure in the exhaust pipe is constant, which is practically the case in an end-pipe measurement. The dilution air flow is controlled using a critical orifice at the pressurized air inlet. As long as the pressure of the sample remains constant the dilution ratio remains constant, in a proportion of about 1:10 [3].
Figure 3: Nozzle detail in the Dekati Diluter. [3].
2.2.2 KTH Ejector diluter
In the previous work, an ejector diluter was designed and built by Jesper Björkstrand, for the KTH purposes. The design of the KTH diluter was based from the basic operation principle of the Dekati diluter.
N1 Nozzle part
N2 Nozzle part
Figure 4: Cross section of the KTH diluter, showing the position X of the straight pipe nozzle
part N1, which moves inside the conical diverging nozzle part N2. [2]
This design was tested in the previous project, concluding that the pressure pulsations were affecting the sample flow, and thus, the dilution ratio obtained. It was one of the main objectives of this project to make the proper modifications in the design of the KTH diluter to make it work properly.
2.2.2.1 KTH diluter modifications
From the results presented in the previous work, it was concluded that this initial design cannot maintain a constant dilution ratio among the range of different engine loads and speeds studied (Engine Speed: 1220 rpm, Engine Torque: from 366 to 1464 Nm) [2]. The possible cause of this was the pressure pulsations that might affect the data collected from the emissions rack [2]. The possible solution leaded to a modification in the design of the KTH diluter, based on the problems detected.
Figure 5: Original KTH diluter nozzle part N2.
The first modification was thus made to the nozzle in the KTH diluter. The straight shaped borders in the nozzle part N2 were changed to rounded shaped borders, as it is shown in Figure 6. This change was meant to allow more pressurized air to get in the mixing chamber, and also to produce the increment of pressure and induce the high speed flow exactly in the beginning of the diverging angle part of the nozzle.
Figure 6: Nozzle part N2 with rounded borders. First modification of the KTH diluter.
The idea was to set the distance X of the N1 part, shown in Figure 4, exactly at the beginning of the diverging angle present in the N2 part. This was practically impossible because the straight pipe of the N1 part was too long for reaching that position.
The second modification was therefore made to the nozzle of the KTH diluter. The length of the straight pipe in the N1 part was reduced from 56 mm to 43 mm. With this change it was possible to set the position X right before the diverging angle of the N2 part, and also it was possible to cover 14 mm from that initial position. The nozzle part N1 with the modification can be seen in figure 7.
Figure 7: Nozzle part N1 with reduced straight pipe. Second modification of the KTH diluter.
After the two modifications made to the KTH diluter, the inlet and outlet flow was measured. This test was made, blocking the excess air outlet and setting the compressed air flow to 2 bar (relative pressure). Two flow meters were used depending on the range of the flow to be measured. The first was the Bios DryCal DC-Lite flow meter, which can measure air flows from 0,02 dm3/min and 20 dm3/min. (20 dm3/min is equal to 20 l/min). The inlet sucked air was measured in a range between 14,4 l/min and 5,6 l/min. The outlet flow was measured with a laminar flow meter that was already installed in the cell bed. It was used instead of the DC-Lite meter because the outlet flow was above the 20 l/min.
Both exhaust inlet and excess air outlet were blocked, so the measured flow was going to be only the outlet from the diluter. From the DC-Lite meter, the outlet flow was 20,5 l/min. From the laminar meter, the outlet flow was 17,5 l/min. So it was established that for this test, it was necessary to add 3 l/min to the measured flows obtained from the laminar meter, which was less accurate.
After this correction, the inlet and outlet flows of the KTH diluter were measured. The main purpose of this flow measurements was to evaluate whether it was possible to get more exhaust gases sucked into the mixing chamber [2].
Inlet and Outlet Flow vs. Position X
0 2 4 6 8 10 12 14 16 8 10 12 14 16 18 20 Position X (mm)
Inlet flow (l/min)
0 5 10 15 20 25 30 35 O u tlet flo w ( l/m in ) Inlet Outlet
Figure 8. Inlet and outlet flow measurements in the KTH diluter.
The maximum inlet and outlet flow occurs when the distance X is set to 10 mm. as shown in Figure 8.
That position coincide with the beginning of the diverging angle in N2 part of the nozzle. This means that the modifications resulted in a significant increase in the flows that the diluter can handle. The modifications increased the maximum flow measured in the inlet (about 600%) and in the outlet (about 46%) from the previous values reported by Björkstrand.
This should allow controlling the exhaust that is getting in the diluter and the pressure will drop before the flow gets into the mixing chamber. In that way, the dilution ratio should be fewer dependant on the pressure pulsations from the engine. This modification is intended for a continuation project.
2.2.3 Dump volume for avoiding pressure pulsations
The results from the previous research concluded that the pressure pulsations were affecting the dilution ratio that could be measured from both KTH and Dekati diluters. In this project, the tests made with the Dekati diluter that will be shown later in the report also revealed a problem in the obtained value of the dilution ratio. From the measurements of the different gases involved (NOx, CO, HC, and CO) there were different values of the dilution ratio calculated from the emissions collected by the rack. The dilution ratio must be same constant value for all gases studied.
Therefore, the assumption that the problem should be pressure pulsations coming from the engine affecting the flow measured, derivate in a proposal for calibrating the dilution ratio. That proposal was to build a dump volume to put it before the probe that lead the flow into the emissions rack. This device should help to avoid the pressure pulsations if these were the cause of the different values of the dilution ratio. Figure 9 shows the built device.
Figure 9: Dump volume built for avoiding pressure pulsations effect on measures.
The dump volume is a cylinder made from steel that has an approximate volume of 2,375 liters. This dump volume can be connected using pipes of 14 mm and 16 mm of diameter, which are the available in the test cell for leading the flow to the emission rack.
2.2.4 Diluter with heating system (Proposal)
This part of the research proposes a solution for trying to avoid the nucleation and condensation problems in the exhaust sample. The idea is to use a two-stage dilution system. In this setup, the first stage is heated. The condensation and nucleation is prevented as the vapor pressures of volatile components are decreased with the heating. Moreover, the secondary dilution is carried out with cold air to cool the sample in a controlled manner, for preventing lost of particles onto the transport lines due to thermophoresis [3].
The dilution air for the first diluter should be heated up to the temperature of the sample gas. However, both dilution ratios should be calibrated separately [3].
Figure 10 represents the measurement setup, with two dekati diluters. [3].
Figure 10: Double dilution system setup. [3].
2.3 Particulate mass measurement devices
In this project, the main objective was to be able to measure the particle mass from the sample extracted from the exhaust flow. In the previous stage, the sample flow was diluted in order to send it to the particle mass measurements devices. The device intended to perform this measurements was the Tapered Element Oscillating Microbalance, or TEOM. These measurements should be compared with a reference value for verifying the accuracy of the findings. Therefore, the measurements would be compared with the values obtained by the smokemeter device, which measure the smoke in the exhaust and therefore the mass concentration in the exhaust pipe from the engine can be calculated.
2.3.1 Tapered element oscillating microbalance (TEOM 1100)
The TEOM 1100 monitor, manufactured by Rupprechet & Patashnick Co., Inc, is a device that performs measurements of the mass of the particles present in the diesel exhaust in real time. It needs a diluted flow that is lead to the inlet (See Figure 11)
Sample Inlet Vacuum pump Inlet Pressure Drop Across filter gauge.
Figure 11: TEOM instrument for measuring particle mass from the exhaust.
The TEOM uses a combination of filter sampling and a resonant frequency detector is used to measure transient particulate mass from the exhaust. The aerosol is pulled through a heated inlet and particles are collected at a filter mounted on the tip of an oscillating hollow quartz rod through which the sample is pulled (See Figure 12). As the oscillating frequency is a function of the temperature, the tapered tube, filter, and sampled air are temperature stabilized at typically 50°C. The heating prevents condensation and provides an standard sample condition removing the semi-volatile components. However, measurements may be affected by the condensation of moisture from the exhaust [1].
Figure 12: TEOM instrument internal components. [1].
The TEOM filters are produced to exactly standards, for maximize the efficiency of the instrument under a wide variety of conditions. The filter should be replaced when the pressure drop across the filter reach the 15 inHg in the pressure drop gauge placed in the instrument. The procedure of changing the filter has to be done with extreme precaution, only touching the filter with the special tool provided [4].
Figure 13: TEOM filter insertion and removal procedure [4].
The TEOM is connected to the engine test bed’s computer. From there the user can control the power switch of the instrument. The frequency can be monitored in the test bed’s computer system. However, to get a proper data from the measurements, some modifications to the electronic input system were made and they are going to be explained in detail in the electrical modification section of this report.
2.3.1.1 Calibration of TEOM
Since the TEOM data processor unit that controls the TEOM cannot log data, the frequency signal has to be logged in the engine test bed along with another data collected simultaneously during tests.
From the formula that gives the relationship between the frequency and the mass in the TEOM instrument, a mathematical correlation was obtained to determine the accumulated mass in the filter. This correlations and the calibration graphic are shown as they were obtained during the previous work (See reference [2]) in combination with this project. There is however a difference in the value of the constant Kt and the value of the equivalent point mass of the tube
m
tube(p). The values differ due to dissimilar calculation methods.m
k
f
π
2
1
=
……Eq. 1: Frequency (f) against mass(m)
The constant k is unknown, and then Eq.1 derivates to Eq. 2
tot
m
Kt
f
=
1
…….Eq. 2Where Kt is the calibration constant and mtot represents the total mass of the
filter in addition to the equivalent point mass of the oscillating tube where the filter is mounted. Thus, it derivates Eq. 3 for the mass.
m
tot= m
filter+ m
tube(p) ...Eq. 3Four filters were set up by adding mass to them in order to weight them on a scale of 10-5 gram accuracy. Each filter was tested into the TEOM and reported a different value of the TEOM frequency.
TEOM Frequency vs Sqroot(1/(Mtube(p)+Mfilter)) y = 83,346x R2 = 0,999 170 180 190 200 210 220 230 240 250 260 270 2 2,2 2,4 2,6 2,8 3 3,2 Sqroot(1/(Mtube(p)+Mfilter)) (1/g^(1/2)) TEO M Frequency ( H z)
Figure 14: TEOM frequency relationship with the filter mass. This graphic will help predicting
the particle mass concentration in the exhaust flow.
2.3.1.2 Electrical modification in TEOM electronic input circuit
The first intent to log the frequency in the engine test bed, reported a scattered sample during all test. This means that the machine working at idle condition did not had a constant value for the frequency of oscillation. Figure 15 shows the initial behavior of the frequency along the test at idle condition, without exhaust from the engine.
TEOM Frequency vs time (TEOM idle, without exhaust)
196,2000 196,2100 196,2200 196,2300 196,2400 196,2500 196,2600 196,2700 196,2800 196,2900 196,3000 196,3100 196,3200 196,3300 196,3400 196,3500 196,3600 196,3700 196,3800 196,3900 196,4000 0 20 40 60 80 100 120 140 160 180 Time (s) TEOM Frequency (Hz)
Figure 15: Initial TEOM signal against time. Scattering in the TEOM signal.
The scattering present in the data collected was a problem to solve in order to get proper measurements with the exhaust. For the analysis of the possible causes of this problem, this issue was consulted with Eng. Mikael Hellgren, expert in mecatronics, and Prof. Hans-Erik Ångstrom, who build the engine test cell at KTH. The conclusion was that there must be an electrical problem affecting the TEOM electronic input circuit that allows the signal to travel from the TEOM device to the engine test bed program that log it during the tests.
Figure 16: Initial TEOM electronic input circuit.
After the analysis of the circuit in Figure 16, it was concluded that the signal was having an infinite amplification due to the configuration of the circuit. Therefore, it was proposed a modification of the circuit to avoid the infinite amplification and make a feedback of the signal to have a limited amplification of 6 times. In that way the amplification of the signal should be controlled and thus, the scattered values obtained while logging the frequency at the engine test cell program should be less. The modification was made between the A and B points shown in Figure 14. A detail view of the electronic modification is showed on Figure 17.
In the modified circuit showed in Figure 17, the amplification is found by the following formula: 1 6 2 10 + = = K K
Amp , so the amplification is effectively limited to a
value of 6 times.
After the electronic modification of the input circuit, the frequency of the TEOM was tested again at idle condition without exhaust. The result of this test is showed on figure 18.
TEOM Frequency vs Time (Teom idle, without exhaust, After circuit modification)
196,2000 196,2100 196,2200 196,2300 196,2400 196,2500 196,2600 196,2700 196,2800 196,2900 196,3000 196,3100 196,3200 196,3300 196,3400 196,3500 196,3600 196,3700 196,3800 196,3900 196,4000 0 20 40 60 80 100 120 140 160 180 Time (s) TEOM Frequency ( H z)
Figure 18: Teom signal against time after circuit modification.
2.3.2 Smokemeter:
The smokemeter is extensively used for steady-state diesel engine calibration. In this device, a certain volume of the exhaust is drawn through a filter, and a photoelectric measuring head measures the blackness of the filter by calculating the reflection of a light beam at a certain angle[1]. This gives the Filter Smoke Number (FSN) or Bosch number, which can be used to calculate the amount of black soot. In this engine test bed is used a smokemeter AVL 415S, and the FSN number will be logged during the TEOM tests to have a point of comparison for the TEOM results for the particle mass concentration.
3. Test procedures
The purpose of the tests were divided in two main objectives. The first was to determine the dilution ratio of the Dekati diluter, and the second was to measure the TEOM frequency in order to get the particulate mass concentration from the engine exhaust.
3.1 Flow dilution tests
For these tests, the engine was operated at selected test points that were obtained from a diesel emissions lab from advanced courses of combustion engines, and there were set to the values shown in table 1.
Table 1: Engine operation selected point for flow dilution tests.
Point Speed (rpm) Load (Nm) Load (% of max)
1 1220 366 20
2 1220 722 40
Figure 19: Torque vs. Engine speed chart from KTH lab. Report, corresponding to Scania
DC1201 diesel engine
The sample flow was taken from the engine exhaust, and diluted with the Dekati diluter. The emissions from NOx, CO and CO2 gases were collected during the tests in
the Boo instruments emission rack installed in the test cell. The dilution ratio was obtained dividing the emissions data collected before dilution by the emissions data collected after dilution, for each gas. All points were let to stabilize along almost 3 minutes before collecting the data, and then were logged for two minutes. From each measure, an average has been calculated and therefore, the results were plotted in graphics that are shown in the results section.
3.2 Particulate mass concentration tests
For these tests, the engine was operated at selected test points obtained for standard emissions calibration of the engine. The points were set to the values shown in Table 2.
Table 2: Engine operation selected point for particle mass concentration tests.
Point Speed (rpm) Load (Nm)
1 1220 458
2 1220 915
3 1220 1373 4 1220 1830
The sample was taken from the engine exhaust, diluted through the Dekati diluter and then sent to the TEOM instrument to log the frequency of oscillation of the tapered element with the filter that collected the particles during tests.
The TEOM instrument has a data processor unit that regulates the sample flow rate and temperature, and it needs about 20 minutes to stabilize those parameters before collecting the data.
Each point was measured for 3 minutes. The data collected was processed and a graphic was obtained for each measure point, of the TEOM frequency against time. With these graphics and the appropriate correlations used in the result section, the particulate mass concentration for each measure point it can be predicted.
Simultaneously a sample from the smokemeter was obtained almost at the end of each measure time to have the FSN number, and also the emissions for NOx, CO, CO2
and HC were logged during the measure time. This was done to be able to calculate the particulate mass concentration from the program developed by Hans-Erik Ångstrom for calculating the particulate mass concentration from the emissions values and the FSN number.
4. Results
In this section, the results from both types of measurements are presented in different sections, correspondingly. First, the findings in the dilution ratio investigation, and then the achievement in the particulate mass measurements during this investigation.
4.1 Flow dilution results
The measurements were made with the Dekati diluter as it was described in the test procedure section. The consideration and proper connections were made, following the Dekati diluter manual. The CO2 in the air was estimated to 0,05%.
4.1.1 Dekati dilution ratio test
Dilution ratio vs. Engine load (Dekati diluter)
8,0 8,5 9,0 9,5 10,0 10,5 11,0 11,5 12,0 12,5 13,0 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 Load (Nm) Dilution ratio NOx CO CO2 CO2 w/bak
Figure 20: Dekati dilution ratio results. Engine operation points from Table 1. The CO2 w/bak
line shows the results for the dilution ratio when the CO2 in the air is substracted from the
emissions rack value.
Figure 20 indicates that the Dekati diluter has a stable dilution ratio of 1:8 for NOx and CO emissions, while it shows a value of about 1:10 for CO2 emissions. The
4.1.2 Dekati dilution ratio with dump volume test
Dilution ratio vs. Engine load (Dekati diluter with dump volume)
8,0 8,5 9,0 9,5 10,0 10,5 11,0 11,5 12,0 12,5 13,0 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 Load (Nm) Dilution ratio NOx CO CO2 CO2 w/bak
Figure 21: Dekati dilution ratio with dump volume results. Engine operation points from Table
1. The CO2 w/bak line shows the results for the dilution ratio when the CO2 in the air is
substracted from the emissions rack value.
4.2 Particulate mass measurements tests
TEOM Frequency vs. Time (1220 rpm, 458 Nm)
196,2770 196,2780 196,2790 196,2800 196,2810 196,2820 196,2830 196,2840 196,2850 196,2860 196,2870 196,2880 196,2890 196,2900 196,2910 196,2920 196,2930 196,2940 196,2950 196,2960 0 20 40 60 80 100 120 140 160 180 Time (s) TEO M Frequency ( H z)
Figure 22: TEOM frequency against time for measure point 1 of Table 2.
TEOM Frequency vs. Time (1220 rpm, 915 Nm)
196,2670 196,2680 196,2690 196,2700 196,2710 196,2720 196,2730 196,2740 196,2750 196,2760 196,2770 196,2780 196,2790 196,2800 196,2810 196,2820 196,2830 196,2840 196,2850 196,2860 196,2870 196,2880 196,2890 196,2900 0 20 40 60 80 100 120 140 160 180 Time (s) TEOM Frequency (Hz)
TEOM Frequency vs. Time (1220 rpm, 1373 Nm) 196,2550 196,2560 196,2570 196,2580 196,2590 196,2600 196,2610 196,2620 196,2630 196,2640 196,2650 196,2660 196,2670 196,2680 196,2690 196,2700 196,2710 196,2720 196,2730 196,2740 196,2750 196,2760 196,2770 0 20 40 60 80 100 120 140 160 180 Time (s) TEO M Frequency ( H z)
Figure 24: TEOM frequency against time for measure point 3 of Table 2.
TEOM Frequency vs. Time (1220 rpm, 1830 Nm)
196,2510 196,2520 196,2530 196,2540 196,2550 196,2560 196,2570 196,2580 196,2590 196,2600 196,2610 196,2620 196,2630 196,2640 196,2650 196,2660 196,2670 196,2680 196,2690 196,2700 196,2710 0 20 40 60 80 100 120 140 160 180 Time (s) TEO M Frequency ( H z)
From the Figures 22, 23, 24 and 25, it can be observed that the tendency of the value of the frequency is decreasing with time, which is logical because the filter is accumulating particles and becoming heavier.
With the help of a trend line, can be established the approximate initial and final values of the TEOM frequency. The TEOM frequency values, and the respective value of the mass of the filter obtained by the Figure 14 correlation, are shown in Table 3.
Table 3: TEOM frequency values and their respective value of the filter mass, from Figure 14
correlation
Point TEOM Frequency (Hz) Filter weight (g)
1220 rpm, 458 Nm 196,2856 0,07876837 196,2808 0,07877730 1220 rpm, 915 Nm 196,2790 0,07878052 196,2706 0,07879588 1220 rpm, 1373 Nm 196,2688 0,07879930 196,2577 0,07881962 1220 rpm, 1830 Nm 196,2609 0,07881370 196,2559 0,07882294
Therefore, a mathematical progression can be made for finding the particulate mass concentration in the exhaust of the engine:
• The difference of the initial and final weight of the filter gives the accumulate weight in the filter of the TEOM.
• That mass divided by the measure time, 180 seconds, gives the mass concentration in the TEOM.
• That mass concentration multiplied by the dilution ratio, in this case obtained nominal by the Dekati data sheet, set to a value of 8.45, gives the mass concentration of the exhaust flow that enters into the diluter.
• With the fuel flow and the air flow, obtained by the log file of the test, it can be obtained the flow of the exhaust pipe of the engine, in grams per second.
• With the pressure after the turbine (P2T) and the temperature after the turbine (T2T2), it can be obtained with ideal gas equation
T R M P ⋅ ⋅ =
ρ and the proper
• With the density it can be obtained the total engine exhaust flow in liters per second, and properly converted to liters per minute.
• Finally, with a linear proportion is obtained the particulate mass concentration in the total exhaust of the engine, in milligrams per second.
The total progression and the results for each measure point are shown on Table 4.
Table 4: Mathematical progression to calculate the particulate mass concentration in the
exhaust of the engine.
Parameter Point 1 Point 2 Point 3 Point 4
Acc W (g) 8,93753E-06 1,5362E-05 2,03183E-05 0,000009238
Acc W (mg) 0,008937529 0,01536203 0,020318333 0,009238415
Time (s) 180 180 180 180
TEOM Mass rate (mg/s) 4,96529E-05 8,5345E-05 0,00011288 5,13245E-05
Dil Ratio 8,45 8,45 8,45 8,45
Mass rate ex. bef dil (mg/s) 0,00041956 0,00072116 0,00095383 0,00043369
Flow rate TEOM (l/min) 3 3 3 3
Fuel flow (g/s) 3,472 6,450 9,575 12,645 Air flow (g/s) 155,33 191,51 238,05 297,17 Exhaust Flow (g/s) 158,80 197,96 247,63 309,81 P2T (bar) 0,991 0,996 0,999 0,991 T2T2 (°C) 285,23 373,26 455,96 497,45 p (g/l) 0,61906 0,53745 0,47792 0,44857
Exhaust flow (l/min) 15391,18 22099,79 31087,69 41440,22
Total Particulate Mass rate
in exhaust by TEOM (mg/s) 2,15 5,31 9,88 5,99
During tests there were found the values of the emissions of NOx, CO and HC, and the necessary parameters for finding the particulate mass rate in the exhaust. The Avgaskemi program version 3, created by Hans-Erik Ångstrom, was used and it will allow to have a parameter of comparisons with the TEOM findings.
Table 5: Parameters and results obtained of the particulate mass concentration in the engine
exhaust, from the Avgaskemi program.
Parameter
Measure points, 1220 rpm at:
458 Nm 915 Nm 1373 Nm 1464 Nm Atoms of CH in fuel (MK1) 1,875 1,875 1,875 1,875 Atoms of O in fuel 0 0 0 0 Engine Power (KW) 58,51 116,9 175,4 233,8 Exhaust Flow (Kg/s) 0,159 0,198 0,248 0,310 Lambda 3,04 2,02 1,69 1,59 HC (PPM) 127 129 127 110 NOx (PPM) 415 667 839 874 CO (PPM) 99 113 230 154 Bosch FSN 0,78 1,07 1,26 0,79 Sulphure % in fuel 0,004 0,004 0,004 0,004
Avkemi Mass rate (mg/s) 4,24
6,44 9,06 7,54
Finally, it was established a comparison between both procedures to obtain the particulate mass concentration in the engine exhaust. The results are shown in Table 6.
Table 6: Comparison of the values of the particulate mass rate obtained by both methods.
Method Point 1 Point 2 Point 3 Point 4
Total Particulate Mass rate in
exhaust by TEOM (mg/s) 2,15 5,31 9,88 5,99
Avkemi Mass rate (mg/s) 4,24 6,44 9,06 7,54
Difference (%) 49 17 9 20
The methods are based on experimental data and have an error in their accuracy. It is notable that both methods give results that are close in their values. This indicates that the method developed in this investigation is a possible alternative to get an acceptable approximation of the particulate mass concentration present in the engine exhaust.
5. Discussion
The results obtained in this investigation lead to a continuous development in this field of research. An alternative method of calculation of the particulate mass concentration in the engine exhaust has been established, and it has to be further developed and corrected to obtain more accurate results in the future.
It is intended for a continuation project, to find the proper method for calibrating the dilution ratio. The device for dumping the flow has to be developed and tested after making the necessary modifications to make it work properly.
The pressure pulsations are still believe to affect the accuracy in the dilution tests, as showed by the different dilution ratios obtained with the Dekati diluter during the tests.
The electronic improvement of the input circuit of the TEOM has been a successful advance, cleaning the signal obtained by reducing the scattering to minimal in comparison with the previous electrical configuration.
In addition, the modification made to the KTH diluter has improved the capacity of operation of the diluter. Before improvement, it was limited to work with an inlet flow of about 2 l/min and an outlet flow of 16 l/min. It has now been improved to manage a inlet suction of 14 l/min and a outlet flow of 30 l/min.
6. Conclusions
It was established an alternative method of predicting the particulate mass concentration in the exhaust flow of a diesel engine. The electrical modification of the inlet circuit has worked successfully and the TEOM signal can be obtained with more precision, and without scattering in the signal.
This alternative method needs to be developed and also improved by making the proper modifications to have an stable dilution ratio in the sample. The sampling system also can be improved by implementing the heating system and thus, reduce the problems due to condensation of water and hydrocarbons.
7. References
1. Bergman, H., T. Benham, O. Berg, A. Kannel, A.M. Rydström, J. Samuelsson and C. de Serves, Survey of dilution and measurement techniques for engine exhaust particles, TRITA-MMK 2003:33, ISSN 1400-1179, KTH, Stockholm (2003).
2. Björkstrand, J., “Design of an ejector diluter and a system for particulate mass measurements of diesel exhaust gases”, Master of Science Thesis MMK 2006:04, KTH, Stockholm (2006)
3. Dekati Ltd. Technical note. “Dekati ejector diluter in exhaust measurements, version 2.1”, 2003.
