Industrial leak testing
From Leakpedia
The industrial leak testing
The leak testing is the industrial proceedings necessary to verify and to measure the pneumatic tightness of the produced components. Particularly this document is made for the technicians employed to the end line testing of serial products. This phase of the industrial process is also called leak test or leakage detection.
Historically the companies that are better equipped in this production phase are those ones producing articles that are essentially very dangerous, or with high technological value. Therefore we can draft the first "generation" of production types where the sensibility of this proceeding has been realized from the beginning:
- Aerospace
- Biomedical
- Navy
- Pharmaceutical
- Gas processing and distribution / burners
- Liquid fuel components
Only at the turn of the eighties and the nineties it has begun a campaign in order to extend these kinds of tests for a larger products range. Around this period, in fact, people become to understand that this kind of test could be extended with a lot of advantages also to those products that were not dangerous in case of leakage. If, for example, small oil leaks in motors or in mechanical transmission were considered acceptable in the seventies and in the eighties, from the nineties a little defect of this kind meant a non conformity by the final customer. Thanks to the application of this principle of technical availability with quite short times and low cost this testing methods expanded, and it continues to expand. Therefore nowadays it is not possible to draft a complete list of companies where this kind of test is considered "necessary", and we can generalize that every product that can contain some gas or some fluids can take advantage of these tests.
Kind of leak tests
For leak test we mean a proceeding that can verify the pneumatic tightness of a piece. We recognize above all two kinds of equipments for leak testing:
- A) verification system, generally controlled by an operator, with location of the leak point:
- Test in water with the piece under pressure (visual inspection)
- Test with soap with the piece under pressure (visual inspection)
- Test with reagents under pressure and UVA lamps (visual inspection)
- Test with gas (Helium) under pressure (control with a mass spectrometer)
- Test with hot air (visual inspection at infrared)
- Test with variation of dielectric on plastic pieces (ion high-voltage system)
- Test with hydrogen gas (localization inspection)
- B) Automatic systems with indication of Good, Waste and value of the leakage:
- Measure by the flow measurement of the piece under pressure.
- Measure by the differential decay of pressure between the piece under test and the sample of reference
- Measure by the pressure decay of the piece under pressure
- “Interception” measure of pressure increasing under bell
If on one side the first class of equipments (A) represents an irreplaceable area test in statistical control and off-line, allowing to find very small leaks and analyzing the fault directly in a visual way, the equipment of kind (B) represents the effective "barrier" or filter of end of line concerning the non-conforming production. The equipment of kind (B) allows also to display during the time any drift of quality, if applied on all the production. Considering the evidence of working principles of measuring systems of kind (A), we shall not remain on a detailed technical description. We have only to consider that the systems based on mass spectrometers (helium “sniffer"), even if expensive both in terms of plant and of management, they are the top for sensitivity in terms of determinable leak compared to any other system described in the present document. On the contrary, the systems with water, soap or reagents, if on one side they allow to detect very small leaks with cheap operating cost, they cannot be automated and then they need necessarily a visual inspection, and in this way an evaluation by an operator. This document wants also to deepen in technical details, the working principles with related considerations concerning the equipment under pressure of the kind Good/Waste.
General principles of automatic leak test equipments
To introduce a detailed description of the several kinds of automatic leak test equipments under pressure, it is necessary to define some characteristics that are common to the different working principles. Every described system has in common the necessity to create a dump or pressure difference between the area considered tight and the outside of this body. This phase is called filling phase. Generally the filling can be made both with positive pressure and with negative one, both with pressure (or depression) applied from the inside or from the outside of the piece under test. After this phase we will have an settling phase, necessary to stabilize the pressure or the flow values of leak measurement. Only at the end of these two phases we will have the execution of the real leak measure, in the different strategies that we will analyze in detail. We report, to be clear, a graph with a typical trend of the pressure during a leak test (pressure decay).
Selection of filling and pressure testing
The strategy of filling, that is if it is from the inside of the piece (more general case) or from the outside (bell), the type of filling and therefore of pressure or depression and the value of this pressurization, must be selected for each case, analyzing the piece to be tested. For this selection, the first parameter we have to consider is the pressure value we have to use to make the filling, therefore the test. Considering the use of common industrial compressed air, this value can be included in more common cases between –1 Bar and 10 Bar, and in case of leak test together with breaking or explosion test this value can arrive also over 40 Bar. In spite of what you can assume intuitively, the use of high pressure values degrades the overall performance of the test, because if on one side the value of the leakage measured is increasing proportionally, that is, however, proportional to the pressure or flow measures, the use of high pressure complicates the trend of the filling phase and of the following settling phases or stabilization ones. Therefore, usually, it is better to prefer tests and filling made at a low pressure (less than 1 bar). The use of a filling in depression can, for example, better the tightness of the piece during the test phases. In case of jars or of pieces with a large "open" section, for example engine oil "pan" or halves of housing, a simple soft rubber base is enough to make the piece tight, without having to use excessive contrast forces. The filling in depression can be distorted in case of tests on pieces of plastic soldered, because the depression tends to make sticky and then to “ glue” the faulty welding. In case of pieces of plastic soldered, the high pressure helps to expand the possible defect, then the test made at a pressure between 3 or 8 bars, can join to a leak test, any test of strength of the soldering. Particular attention should be placed when the piece to be tested is composed by “non-linear” tight mechanism such as valves or leaf springs and tests have to be made or at a much lower or much higher pressure than the operating point of these valves.
On mechanical pieces such as cast iron motor housing or of motorcycle transmissions you have always to consider the presence or the absence of seals or of any components guaranteed up to a known pressure. The specifications of tests for pieces for gas and kitchen indicate the leak at low pressures, usually 150 mBar. A very important note is for those metal pieces that are used to expand after filling phase under pressure. These pieces, such as windings, heat exchanger etc, tend to expand mechanically at the end of the filling phase, gradually according to the introduced pressure. Considering the cases in which these elements have to be tested at a quite high pressure necessarily (4…7 Bar), and where the expansion cannot be compensated by the settling phase, only with time of prohibitive length, the use of a pre-filling at a higher value than that one of the test itself allows to obtain excellent results of expansion/relaxation, limiting the total testing time drastically.
Synthetically, the choice of the pressure value at which you have to make the test, has to be similar from one side to the real working pressure of the component, considering each time the pros and cons of different pressure levels. The choice of a filling in “bell”, and therefore form the outside of the piece, usually in depression, is discussed here below in its proper paragraph. The gas used is common to every kind of filling and it is in most cases at compressed air. This air has to be intended as a filtered one, obviously without any oil, and dehumidified.
In case you use industrial air of a general purpose circuit, the application of a cylinder or of a local expansion vessel to the leak test equipment will better the characteristics of temperature changing between the air and the piece itself. Instead of the air you can use some gas with smaller atoms, such as helium, because they increase the leak fluidity and they emphasize the sensitivity of the test. Finally it is necessary to consider the use of inert gas such as nitrogen, in case of tests in components already treated with explosive or flammable elements, such as tests on car gasoline branches or on components for fuels generally.
Procedures modes of the filling phase
Until now, we have analyzed this filling phase without considering how the equipment can do this. The strategies that we know and that can be applicable are substantially two: the level filling and the time filling.
Level filling
Considering that this phase essentially consists in bringing gas through electrically controlled valves, the first method is to drive the valve until reaching the desired pressure, or a little bit higher value in order to compensate the pressure reduction in the settling phase. The second strategy is to drive this entry for a fixed time, checking the pressure value reached only at the end, in order to verify its acceptability. Analyzing the typical development of a complete test cycle (Graph1) it is necessary to understand that the total repeatability of the tests depends on the constancy of repeatability of the different phases of test cycle. The test cycle must begin always from a stabilized piece (both thermally and mechanically), and then from an environmental pressure; the filling must be at a pressure value as constant as possible over time, and obviously the duration of adjustment and test phases have to be strictly constant. Over all you have to consider that the test phase will contain a spurious decay value usually not depending on the parasitic leak, but aver all on a settling phase ending still running; therefore, and particularly during tests at the limit of due time, the change of duration of the settling phase involves great mistakes of total repeatability of the test. Considering what said before, the level filling improves because in addition to eliminate unnecessary downtime, it ensures the starting of the settling time from a known pressure value, that is that one of the filling pressure just reached.
Time Filling
In some conditions the strategy described above cannot be used. In case you have to test elastic pieces such as rubber connections, or feeding bags or medical one, the level filling cannot be used because it produces wrong fillings caused by expansions, and for this reason the consequent pressure decay, at the end of filling phase. In this case you can use the constant “time” filling strategy, taking care to check the pressure and the flow of the filling, in order to reduce the mistakes caused by the variances of settling time.
At similar cases you can relate the filling of complex pieces (ex. Complete engine, very complex gears) that is where the air that you introduce in a primary volume has the necessity of a ime in order to fill the secondary volumes connected by interstices having an air passage smaller than the total one of filling phase. Practical experiences demonstrate that particular settings of the filling phases, based on pressure regulators controlled by a management software of the equipments, optimize the time and the repeatability of these fillings. These strategies use pressure “ramps” for the filling.
Settling phase
Excluding “bell” tests and some kinds of time filling tests, all the described systems have the necessity of an adjustment phase to be done at the end of the filling. This time, strictly constant and repetitive is necessary to reduce and even to eliminate the effects of adiabatic heating and the turbulence of the filling phase, submitted to the force of pressure and to the volume variance caused by the movement of the filling valve in case of tests on little volume.
In this phase the equipment does not make particular functions: the necessary tests are the ones concerning the absolute value of the introduced pressure, that has not to decrease under the tolerance limits according to the filling value.
A minimum limit of pressure can indicate that there are great leaks on the piece, in order to reduce the total testing time. If you compare pressure trend tables stocked during previous tests, or predictive tables concerning pressure trend, you can better a lot this kind of test (great leakage) in this particular phase.
As we have said for the filling phase, it does not exist a constant rule to define the duration of the adjustment phase, that has to be determined by attempts practically, considering also the worst cases of temperature variation or of possible difference of elasticity of the piece to be tested in different production lots.
Leak measure test of the piece under pressure by flow test
This system allows to measure in a direct way the flow or airflow made by the leak. At the end of filling and adjustment phases testing time is the time that is necessary to obtain a stable measure of this flow, that is usually very short (ex: 100…300 mSeconds).
As outlined in the Fig.1 the measure of this flow depends on the differential transducer that can read the pressure drop across a charge leak. In order to reduce the quadratic trend caused by the turbulent motion of gas particles, it is used a laminar element that can make linear, in part, this function. (∆p/Flow). For any further information we can also refer to the CNR-UNI 10023. Alternatively to the flow measure made in a volumetric way (precisely with ∆p measure), in the last ten years it has become much more used the mass flow meter application, for example the thermal system or flow meters (“hot-wire” system), because they are much more precise, constant in time, easily available in several scales, and less sensible to thermal variation of measuring gas. Considered as a “historic system” of leak test, the leak measures made with this principle, stand for the following benefits:
- Continuous measure of the leak
- Duration of the testing phase practically zero
- Leak indication in volumetric units (CC/time)
This aspect is the real reason because this principle is yet applied in industrial field. In a natural way, that is without any artifice, with this system it is possible to analyze for a fixed time the leakage, allowing the operator to find it and to repair it in real time during the measure time.
As we have already said, the flow measure being a continuous type of measure, it allows to eliminate of a real time of test timing. This concept, as we can see after, has to be considered in a strictly theoretic way, because if in decay or ∆p systems the adjustment or testing phases can be partly superimposed, in this method, the measure must be made necessarily in the best adjustment conditions.
This characteristic has to be kept between the benefits, even if we will analyze after, some systems that can execute the same measure in a more precise and sure way. Vice versa, if compared with other systems, this principle presents some disadvantages; the first and the most evident one, arises from the complexity and instability of the flow measure. In fact, in addition to the cost of a double measure (pressure and flow) and then to a double control in order to obtain to total validation of the measure, the laminar element that is practically like a capillary one, it is hardly influenced by dirty or distortion. Therefore the measure must be constantly verified with reference nozzles, that presenting themselves as micro pores on a ceramic or metallic base tend to deteriorate, and then they have a limited duration. Moreover, with particular reference to the sketch of Fig 1, any parasite leakage before the element of flow measure can false and mask the possible leak of the testing piece. Therefore, this pneumatic circuit cannot be considered totally at “positive safety”, and it must be verified constantly. in the end the sensitivity of the measure is limited from the scale of the flow meter, while in decay or ∆p systems this limit, even present, it can however be mediated by lengthening the testing time.
Vice versa, if compared with other systems, this principle presents some disadvantages; the first and the most evident one, arises from the complexity and instability of the flow measure. In fact, in addition to the cost of a double measure (pressure and flow) and then to a double control in order to obtain to total validation of the measure, the laminar element that is practically like a capillary one, it is hardly influenced by dirty or distortion. Therefore the measure must be constantly verified with reference nozzles, that presenting themselves as micro pores on a ceramic or metallic base tend to deteriorate, and then they have a limited duration. Moreover, with particular reference to the sketch of Fig 1, any parasite leakage before the element of flow measure can false and mask the possible leak of the testing piece. Therefore, this pneumatic circuit cannot be considered totally at “positive safety”, and it must be verified constantly. in the end the sensitivity of the measure is limited from the scale of the flow meter, while in decay or ∆p systems this limit, even present, it can however be mediated by lengthening the testing time.
Measurement by differential pressure decay between a piece under test and the reference sample
Leak measuring by the differential pneumatic circuit, as in Fig.2, represented till 80's the most ingenious pneumatic artifice in this field in order to put a remedy to the poor precision of sections of measure and electronic acquisition available until then. The system provided a double branch: from one side the piece to be tested. Form the other one an identical piece, but a tight one. Practically, analyzing the pattern, tests developed with the following logic:
The filling phase was made commanding the opening of both valves; The settling phase was done with the valve B closed, and the valve A opened, in order to stabilize and standardize pressure conditions into two branches.
At the end of the settling phase, all the valves were closed. If we imagine the pressure transducer as a membrane (but the experience can also be done with a simple double nozzle mercury) we will have that at the equilibrium the differential pressure is nothing. The possible decrease of the piece under test moves the zero of this measure allowing a very sensible indication of this deviation.
On this principle it is possible to make an easy electrical amplification of the signal coming from the transducer, and to visualize it on a needle equipment with central zero. So with this artifice it was possible to analyze a value of typical decay of 1/50.000 (transducers allowing) at the value of the filling pressure, while the electronic of those times, if it was applied in a pressure gauge system it did not allow to overcome the ratio of 1 / 10.000.
It is clear that the limit was only that one of the electronic measure in terms of resolution and of noise, because the working conditions of the transducers of measure has in any cases the limits of a single metric system. However this transducer had to be rated for the maximum filling pressure, because in case of leakage of the testing piece, the membrane was solicited by the total pressure.
The pneumatics realized in this way presented different disadvantages:
- The comparison examines a tight reference: if this reference has a leakage it coincides with a “masking” of the real leakage measure of the piece under test. This defect was reparable by a continuous check of the system in use by a “good” sample and electric calibration of the measure of “Zero”. This first point classifies this pneumatics not at “positive safety”.
- Difficult in calibration of the differential transducer measure, that had to be executed with a particular verification procedure.
- The measure of the leak that has been made appears to be a measure that does not indicate the real leak of the tested piece, but the relative difference with the sample.
- This does not necessarily coincide with the concept that the reference sample can leak (that also corresponds to reality). For example, if you consider that in the practical use of these systems we will have that the sample piece is mechanically stressed at every testing cycle, while the piece under test only during the phase of its own test. Practically you will analyze a trend of the measured decay that progresses with the hours of use of the system, index of the progressive mechanical adjustment of the reference sample that does not coincide with the adjustment of testing pieces.
Pressure decay measure of the piece under pressure
Leak measure system with the pressure system, provides during the testing time, the measure of the pressure decay inside the piece under test. As reported in Fig.3, pneumatics is reduced essentially to a filling valve and to a measure transducer. If we analyze the layout we will find that any pneumatic defect is due to a leak, then to a waste indication. For this reason the system has been defined at “positive safety”. The only element of risk in this pneumatic circuit is represented by a possible leakage of the filling valve. This problem that is present in many of the pneumatic schemes that we have discussed, can be easily avoid thanks to a particular game valves that replace the simple valve signed in the scheme, and thanks to particular software diagnostic. Opposite, the overall accuracy of this kind of equipment depends essentially on the precision of the measure section (transducer) and on the electronic section of acquisition.
Practically the elements that come to limit the precision ate the electrical noise of the circuit and that mechanical one of the transducer, that correspond to the resolution or maximal number of points within which the full scale measure is decomposed: a system that can guarantee 100.000 points, on a full scale of (for ex) 1Bar, means that you can guarantee a measure resolution of a hundredth of milliBar. Much higher it is this resolution parameter and shorter it will be the time that is necessary fir the decay measure; this means on one side a time reduction –testing cycle, but overall to contain the mistakes due to thermal variations of the gas put inside the piece. The electronic strategies to obtain these results are of different types: they go from the transducer piloting with alternate tensions and the use of refined AD converters to the use of proper zero tracking circuits and measure in windows, but overall a high filtration both electrical and mathematical of the measure.
Particular attention must be paid in determining the point of “zero” of the decay, and practically in the measure of the pressure at the beginning of the testing phase. ”Interception” measure at pressure increasing.
Leak testing with interception
For “interception” leak testing system we mean a system that can detect leaks outside chambers to be tested. The most popular practical and clear example is that one of measure of the leak of valves: you put the air on one side, and you detect the leak on the opposite side. This concept can be extended to every kind of component or piece to be tested considering the possibility to enclose the piece in a tight box called properly “bell”.
The bell system can be used for the applications of interception type, when it is possible to put the piece under pressure from the inside and at the same time to enclose it from the outside. When this is not possible (put the pressure from the inside of the piece) the use of tight bells is however an excellent way to better performances of tests with the conventional pressure system on pieces of a great volume, taking care to create a inter space volume between the piece under test and the bell as small as possible.
Only when you have to make tests at maximum 1 Bar, the use of a system in depression under bell joins to the auto closing function, and then to the box tight, the regeneration of the case usually real of positive pressure inside the piece. Differently from what we described for other methods, the interception system does not need any adjustment time, and any testing time, because it is superimposed to the filling phase making the system faster. Practically, the test starts putting under pressure the testing side of the piece, and during the same time the possible pressure increase cause by leak on the opposite side of the chamber, will be analyzed. This kind of equipments is usually equipped of two transducers of pressure measure; one for the measure of filling pressure, and the other one for the leak pressure detection. Both the transducers have to be sized for the higher filling pressure, in order to prevent the condition of great leak and then of peak pressure on the side of leak detection.
Leak analysis is then the relation between the two pressures and it can be represented as the percentage relation between the two values (leak pressure/filling pressure) or it can be calculated as leak pressure at a nominal value of filling, or in the most sophisticated cases, knowing the external volume or of the bell, given as volume index (CC/hour). From an electronic point of view it is important to size the acquisition considering that tests last generally few seconds or less and that the calculation of the relation is done during the increasing of the filling pressure; so the two measures have to be done in phase with each other and with a sufficiently high frequency in order to avoid mistakes introduction.
Leakage measure units
Leak value can be expressed in two different measure units, that is in a volumetric way (for example CC/minutes) knowing the pressure value at which the leak is referred, or in a pressure way (for example mBar/second), knowing the value of the testing piece. It does not exist any real rule to prefer one scale to another one.
For convenience, when the pieces under test are components for air or gas, if you express the leak in a volumetric way, you will avoid any further procedures of calculation in comparison to the leak limits fixed by any specifications or supply demands. But in most cases, that is, when the piece under test is made to contain no gas fluids, the volume index does not represent in a direct way the real fluid. You have to consider that the equivalence between the two systems is obtained with the following calculation:
In the application of this calculation you have to consider the total volume of the piece, including also tubes and joints volume, and any possible “dead” volume of the equipment;
Moreover, considering that the flow of the leak volume is proportional to the pressure of the gas inside the piece, and that it decrease during time because of the leakage, it is necessary to consider that the mistake that has been introduced (otherwise it can be eliminated by vector calculation) is negligible for little pressure leaks, for example less than 1% of the filling pressure value. The time unit must be coherent between the two characteristics (ex seconds or minutes) and the decay so calculated should be converted in the unit selected by the equipment measure (for example 1 Atm = 1013,25 milliBar). Thanks to the application of microprocessor logic, this calculation is often made by the equipment in order to give an answer of volumetric type. In this case the volume of the testing piece is given in a parametric form, or, in best cases it is calculated by a capacitance during the filling phase. A widespread system and also easily applicable is that one to make the filling of the testing piece by a known volume, controlling the pressure before and after this partial discharge. As we have already said in the description of the flow measure system, in the selection of the unit is however necessary to consider also the problems concerning the validation and the calibration of these measures: in case of pressure decay measure, a simple pressure gauge certified SIT ( the Italian corporation of validation of physical measures) allows this checking.
On the contrary, in the use of leak volumetric measures, besides this validation or pressure calibration, it is necessary a test on flow measure, made with the most traditional method by reference nozzles, that being passive elements, in turn, they should have an independent validation.
In the application of this calculation you have to consider the total volume of the piece, including also tubes and joints volume, and any possible “dead” volume of the equipment;
Moreover, considering that the flow of the leak volume is proportional to the pressure of the gas inside the piece, and that it decrease during time because of the leakage, it is necessary to consider that the mistake that has been introduced (otherwise it can be eliminated by vector calculation) is negligible for little pressure leaks, for example less than 1% of the filling pressure value. The time unit must be coherent between the two characteristics (ex seconds or minutes) and the decay so calculated should be converted in the unit selected by the equipment measure (for example 1 Atm = 1013,25 milliBar). Thanks to the application of microprocessor logic, this calculation is often made by the equipment in order to give an answer of volumetric type. In this case the volume of the testing piece is given in a parametric form, or, in best cases it is calculated by a capacitance during the filling phase. A widespread system and also easily applicable is that one to make the filling of the testing piece by a known volume, controlling the pressure before and after this partial discharge. As we have already said in the description of the flow measure system, in the selection of the unit is however necessary to consider also the problems concerning the validation and the calibration of these measures: in case of pressure decay measure, a simple pressure gauge certified SIT ( the Italian corporation of validation of physical measures) allows this checking.
On the contrary, in the use of leak volumetric measures, besides this validation or pressure calibration, it is necessary a test on flow measure, made with the most traditional method by reference nozzles, that being passive elements, in turn, they should have an independent validation.
Leak evaluation
The definition of the waste, that is when a piece cannot be accepted as a good one during the production, is a parameter that has be defined with care. The ideal case is the application of rules, or of customer specifications, that indicate this limit of acceptability. In this case you proceed to convert, if necessary, this value in the unit of work of the equipment. When we do not have this datum, then the best compromise is that one to proceed to analyze pieces that have defects and stored as archive of waste cases.
However this procedure has two disadvantages:
- The first one is that in some circumstances, for example for soft plastic pieces, the leak can change during the time, and according to the number of tests that you make.
- The second problem you can find in this application of this empirical system, is that, before, it is always necessary to define the filling and the adjustment parameters on a similar piece that is ok for you, and only after this you can analyze the waste.
A suggested method, that even if it is a very empirical and not a very precise one it allows to evaluate in a practical and quick way the leaks, consists in analyzing the wastes in water, starting only from the definition of the filling pressure. Done that, you can find a piece with the less leak value, expressed in bubbles (for ex.) per minute. The warning in this phase is to place the piece in order that the bubbles that will appear, can detach from the body and that they are clearly visible by the operator. Therefore you try to determine the total volume of the leak, given by the sum of the volume in CC of air bubbles, counted and measured visually on the surface of the water. This calculation gives the idea of the volumetric value (CC/minute) and it has to be considered has a departure value for the equipment setting, as well as a value of evidence to test afterwards the possible changes of this leak. A similar system, but that can be applied only on little pieces is to enclose in a graduated flask the piece under pressure, to put the flask with the closed part upward and to put all in the water, in order to leave some air on the top of the container. This air volume increasing, that you can quantify by the scores printed on the flask, corresponds to the quantity of air lost by the piece. If the calculation of the “bubbles” introduces a lot of mistakes caused by mistake (cubic) of the measure of the bubble ray intended for the calculation of the sphere volume, in the system with the flask it is necessary to consider little increasing of leak volume in order to maintain constant the final pressure inside this air volume.
The most complicated case, is that one in which the leak is defined, even in a volumetric way, and it is not concerning gas or compressed air, but other fluids that are used really in the piece. An example can be the case of motors and gears in food sector, lubricated with vegetable oil, and where there is the maximum allowable amount of leak over time of this oil. In this case, it is not sufficient to rely on the theoretical relationship between oil and air flow, because in case of real working we have to consider the different physical conditions of this oil, that is beside temperature and pressure, also the possible decay with time of fluidity (i.e. sludge). We can go on applying nozzles on at least three pieces on which you can ensure the tightness, and making practical tests in order to detect in time leak values of this oil.
Making sure you have used nozzles pre calibrated in air, with three flows different between them it is possible after this test cycle to determine by flask the oil leak. Repeating and putting in a graph the values we have obtained several times (taking care of changing nozzles each time) we can underline the repeatability of the measures and conclude a value close to the leak in the air this recreated.
Practical installation of testing parameters in the system
The installation of leak testing system must be done knowing both the leak and the pressure value you have to make the tests. After having set these data, the first step is, therefore, to set an settling time even an extreme one, in order to make a first check. This check, that must be made on a good sample strictly, is meant to clarify if all works correctly in the whole: plugs or piece connections tightness, control of the mechanic static during the test, control of the thermal variance trend between the piece under test and the gas used for the filling. Then you must verify that this good sample define a decay near to zero or at least less than one third than the programmed one. In this test the decay should never be a negative one (pressure increase). At this point you begin to reduce the settling time, by attempts. In case of equipments with flow measures, you must be content with a settling time that permit a measure of the “good” piece closer to zero in practice. In case of equipments with a differential pressure measure you have to consider what already explained in the concerning section, and that is that on metal pieces of medium and great volume the control of the settling time should happen in more cycles and on long times, taking care to maintain the same reference piece, and to change the “good” pieces under test in order to analyze constancy over time of the settling phase.
In pressure systems, the testing time can be partly superimposed on the adjustment time. Practically the reduction of the adjustment time will be accepted till when the decay value found is less than 50% of the theoretical one.
At this point you can increase the programmed leak value recalculated as the sum between the previous theoretical value and the leak result of the good piece. The theoretical leak value must be intended as a leak limit compared with a good one that has a sure tightness: the parasite decay must not be intended, in fact, as a leakage, but it can be referred to an adjustment decay measure. Repeating several times this operation, and taking care not to exceed the 50% of the decay on good pieces, you obtain the practical calibration of testing parameters of the equipment.
New method to determine loss rates
Zero leakage does not exist, and even if it had existed, it would not have been measurable. For this reason, and also for the technological complexity in the measurement of leakage, it has been historically searching a solution to value the acceptable leak of pieces (as reported in our database). During these years, quality departments have been trying to establish regulations, technical documents or specifications, with the aim of defining the value of allowed leakage. This leakage is usually measured by instruments ATE (Automatic Test Equipment), which typically work in air. Most of the time these documents are not based on physical or mechanical concepts, anyway they assure a total effectiveness thanks their historical application. For example, if a famous car producer tests the radiator with a certain value of leakage in air, using a particular sensitivity of leakage, and during the years he did not find substantial problems in the real quality, it means that the specifications mentioned above practically work. As a consequence, a new necessity arises: to find out a piece alike the one to be tested, and subsequently to obtain a value similar to what historically works. The algorithm supplied by Leakpedia is the solution at this problem. Our analyzer (not reported in documentation, but available) took out the coefficients “industrial sector” and “kind of fluid” from the database of Leakpedia. Who wrote the specifications we have assembled in our database (different and not connected sources) found values of leakage after many attempts. Since the first software analysis it was clear that these values coincide with the ones elaborated by our software. Finally, on the one hand we consider the extraction of coefficients made by our software reliable and effective, on the other hand it is clear that, the more you help us to enrich our database, the more it will supply refined results. The algorithm (as reported here) has totally been our idea, but we think that, already in this first version, it allows manipulating data and obtaining results coherent with specifications collected.
Calibrations and Verifications
We have to distinguish conceptually the periodic calibration from the ordinary verifications into two different processes: For calibration we intend a proceeding that has to verify the total working of the equipment in terms of correspondence with the declared precision limits of the electronic measure and of the pneumatic working. The ordinary verification is done at predefined intervals and it is intended to control the equipment within the limits of the normal use, and then to verify the indication of Good and Waste applying respectively a sample for this hermetic test and another one with a known leakage. Both the proceedings must be done at pre-established time intervals. Analyzing the different specifications or norms concerning these proceedings it comes out that no data is universally applicable. The typical range for the calibration operation can be estimated in 6 or 12 months. The typical range for the operation of ordinary control can be determined by the number of pieces produced and approximately at ranges at 25% of daily production. By consequence, and referring especially to ordinary controls, it is necessary to define a practice of methods and the times in order to validate samples to make these controls. As we cannot report contents of proceeding we received, because they are property of private organizations, and companies, we report only in our opinion an aspect that is contradictory and in common with these proceedings. This aspect is concerning the execution of a leak element (nozzle) to connect with a branch to a hermetic piece, in order to make proceedings of ordinary verification.
By the several researches we made it results that at the date we draw up this document, the Italian organization of validation of physical measures (SIT, that we contacted many time at the numbers of Politecnico of Torino) has not yet made any rule concerning the validation of flow measures of leaks in air or gas. Therefore, and in order not to give any wrong data, we will only treat about our direct experiences of validation, omitting the equipments with measure in flow and focusing only on the equipments with pressure measure. In order to direct the reader of these measures of volumetric type, we advise to refer at the UNI EN 161. The pressure decay systems are the most practical ones in the execution of these phases, because putting simply in scale the pressure measure by a certified pressure gauge you obtain the periodic calibration. This calibration is necessary essentially to set the zero and the full scale of the absolute measure of pressure, and where requested and however only in terms of verification, the decay measure during the test. It is the same we discuss in order to make a sample having a controlled leak.
Measure units conversion table
| Kilo Pascals | mm Hg | millibars | Inches H20 | PSI | |
| 1 atm | 101.325 | 760.000 | 1013.25 | 406.795 | 14.6960 |
| 1 kiloPascal | 1.00000 | 7.50062 | 10.0000 | 4.01475 | 0.145038 |
| 1 mm Hg | 0.133322 | 1.00000 | 1.33322 | 0.535257 | 0.0193368 |
| 1 millibar | 0.100000 | 0.750062 | 1.00000 | 0.401475 | 0.0145038 |
| 1 inch H20 | 0.249081 | 1.86826 | 2.49081 | 1.00000 | 0.0361 |
| 1 PSI | 6.89473 | 51.7148 | 68.9473 | 27.6807 | 1.00000 |
| 1 hectoPascal | 0.100000 | 0.75006 | 1.00000 | 0.401475 | 0.0145038 |
| 1 cm H20 | 0.09806 | 0.7355 | 9.8 x 10^-7 | 0.3937 | 0.014223 |
See also
External links
