The Way
Fowlmere
Cambridge
UK
SG8 7UJ
| Ion Medicare Ltd The Way Fowlmere Cambridge UK SG8 7UJ |
| ||||||||||||||||||||||||
|
Single Orifice (Virtual Leak Diameters VLD) Calibration StandardsOverview: Calibrating leak detectors and moreover the setting of standards remains a subjective issue, with pass/fail standards largely based on traditional practice and individuals preferences. It is not uncommon to discover the same component being produced by different suppliers is tested to very different standards. Setting leak standards has changed little from the 'bucket of water' leak testing where parts known to leak were subjected to an underwater test and the bubbles counted. For example 2 bubbles/sec was equal to 0.6ml/min of air leakage. This leads to over-specifying seeking a wide safety margin. A greater concern is that acceptance levels are set based on what is achievable from the test equipment and testing conditions. This leads to parts being passed or rejected against standards which are not based on the requirement. This leads to additional expense and uncertainty as parts which are 'fit for purpose' end up rejected. One customer has established that an incorrect testing standard resulted in some $128,000 expense over 3 years in manually sealing parts which were totally acceptable for the requirement. Establishing a standard based on the requirement and supported by the logic of engineering, physics and gas laws is the technical correct option. Current calibration methods: Calibrated leaks, come in a range of designs. Most state a maximum allowable volume loss (typically air) at a given pressure, over time. For example a 'standard' may state: At an applied internal pressure of 1bar the component shall be considered rejected if 1ml of air is lost/min. Sometimes standards are written around allowable pressure loss over time, which is very confusing due to internal volume variations in fixtures etc. To achieve a volume loss over time a 'leak' is fitted to the test circuit, or best fitted directly to a component under test. Traditional 'leak' manufacturing involves using a sintered metallic pellet or powder. Basically a small punch is used to compress the sinter whilst a gas (usually air) is applied at a specified pressure. Compression continues in stages until the flow has been restricted enough to meet the requirement, for example 1ml/min at an applied pressure of 1bar. Technically this process is unsound, and the lower the flow the greater the uncertainty. As a compressed material, 'Young's Modulus', (the recovery of compressed materials), is an important factor leading to changes in leak (flow) rates after compression. Furthermore this technique has not proved capable of producing reliable leaks smaller than 10 E-3ml/sec. In addition as a sinter having microscopic pores they are easily reduced or blocked from microscopic contamination in filtered air, oil or moisture. DANGER! The effect of a slowly reducing calibration standard is very damaging as the setter when 'calibrating' will reduce the 'reject level' to that corresponding to the 'standard'. This leads to a gradual reduction in actual 'leak rate' which results in instability and false rejects. Four years ago we became involved with FORD in the testing of fuel containing components to meet emerging USA environmental standards. SHED (Sealed Heated Emission Determination) and the corresponding European standard Zero Evap CARB LEV-11. It is not uncommon to find a drawing specification "zero leaks" Components such as plastic fuel tanks where permeation of fuel vapour through the plastic occurs setting a "zero" standards is as impossible as setting a "zero" mechanical tolerance, it is not possible. Method: Virtual Leak Diameters (VLD) Based on Poiseuilles equation. The presumption of this approach is that an individual leak larger than a given size can be screened out and rejected by the leak test. In real situations, leaks will have different shapes and profiles and components may have two or more leaks or be porous. This means that there are a large number of possible combinations and it is unrealistic to specify a leak rate limit for each case. In addition, the specified leak test method might not apply to every component design. If we assume that a component can only have a single leak and that leak is circular in cross-section, we can define the leak specification based on the size of this hole. As this leak is virtual, we can define other parameters for simplicity. For example, the length of the channel is defined as 1 mm and is not related to the real material thickness, which may well be a casting of some mm thickness. In addition the walls of the virtual leak are smooth and perfect Poiseuilles flow of air, helium etc occurs. For this virtual leak, as defined above, a Virtual Leak Diameter (VLD) can be defined. As the VLD conforms to Poiseuilles flow equation, we can use the following equation to determine the specific mass flow rate for the chosen test method.
Where:
Example To detect leaks in carbon canisters (reclaim filters) using the helium Mass' Spec' leak testing method. If the leak test method is conducted at room temperature (293 K) and the applied internal pressure is 10 kPa (above vacuum) using helium, and the leak is to 1 kPa vacuum, what is the helium VLD specification? The VLD limit from Table over page 1 is 20.0 m. P1 is 10000 Pa and P0 is 1000 Pa.
For helium Therefore the helium mass flow rate dM/dt = 1.58 x 10-11 kg/s As helium has a density of 0.17 kgm-3 (at STP), the helium volumetric flow rate is 9.3 x 10 -11 m3/s or 9.3 x 10-5 cc/s. HELP US TO HELP YOU Email to info@ionmedicare.com
|
|||||||||||||||||||||||
| ||||||||||||||||||||||||
