How Does “Automatic Flow Compensation” Work?

We are often asked questions  about “Automatic Flow Compensation” and how it works.  Automatic flow compensation was designed into MIG gas delivery systems since the  introduction of MIG welding in the 1950's.  It maintains the desired preset flow when spatter builds in the nozzle, torch cables or gas delivery hose is bent and twisted etc.  The following Q&A provides an explanation of why it works:

Question:  With no apparent feedback circuits or moving parts;  how does "Automatic Flow Compensation" work?

Answer: The needle valve or fixed orifice is controlling the gas flow and indirectly the downstream pressure.   This occurs as long as the downstream pressure  is less than about ½ the upstream pressure (actually 2.1 times to be exact, but 2:1 provides sufficient accuracy.)  This is called critical flow or choked flow.  Note all pressures are measured as absolute pressure which is gauge pressure plus 14.7 psi (15 psi is sufficient  for our purposes.)

For a typical MIG system the pressure needed to flow gas through a feeder and torch will be about 5 psi or 20 psia (absolute pressure.)  Therefore the upstream pressure to achieve critical flow will need to be twice that or 40 psia.  40 psia stated as gauge pressure is 40 psia - 15 psi= 25 psi.  Therefore 25 psi is the minimum regulator or pipeline pressure needed upstream of the flow control orifice or needle valve.

Question:  Intuitively, both the upstream and downstream pressure  should control flow why when downstream pressure is less than 2.1 times upstream pressure does that no longer apply?

ANSWER:  It is correct that until the upstream pressure is 2.1 times the downstream pressure both pressures determine flow.  The flow equations are rather complex but flow rate can be calculated given both pressures and orifice size. 

However the gas velocity in the needle valve or fixed orifice can not be greater than the speed of sound (for that gas at that pressure.)  Therefore, for a given upstream pressure, once the gas velocity reaches the speed of sound in the small gas passage a further reduction in downstream pressure has no influence on flow.  Using flow equations the downstream  pressure that causes this "choked flow" to exist occurs when it reaches 1/(2.1) = 0.48 times the upstream pressure (both measured as absolute pressure, i.e. gauge reading + 14.7 psi at sea level. )  

Question:  Why is velocity limited to the speed of sound?

Answer: I recall an explanation given many years ago by my Fluid Flow Professor that may help  explain this phenomena.  When asked why the flow doesn't increase with a continued reduction in downstream pressure?  He said "It would if it knew the pressure was lower but the pressure wave telling it the pressure is reduced can only travel at the speed of sound so it never gets there to let it know!"  

Remember the pressure wave causing  sound only travels at a limited speed.  That is why you see lightening before you hear the resulting thunder!

Question:    How does "Automatic Flow Compensation" control the downstream pressure?

Answer: The drop in pressure from pipeline or regulator pressure to achieve the flow  rate of shielding gas desired is controlled by the needle valve or orifice.  The pressure drops as gas flows through these very small passages.

When choked flow exists, the downstream pressure is controlled automatically by the fact that flow rate is being established in the needle valve or fixed orifice.  The downstream pressure will be whatever it takes to flow the gas coming through the orifice as long as the downstream pressure is less than about ½ the upstream pressure. The orifice flow is controlled only by the pressure upstream in that situation. Therefore as restrictions occur in the system, due to spatter build-up, twisted torch cables etc, the pressure will automatically raise to delivery that flow (or automatically fall when spatter is removed or torch cables have less twist.)

Question:  If the speed of sound in the orifice determines flow, why does flow rate change with a change in upstream pressure?

Answer:  The velocity will still be the speed of sound  (at a higher pressure the speed of sound will increase.)  The volume of gas will also increase because the density of the gas increases with pressure.  Velocity is  still the speed of sound (at the higher pressure)  but the volume of gas measured at Standard Pressure and Temperature (that is what is set in welding procedures and what is paid for) increases.

Question:  What happens when the pressure upstream of the orifice is  less than 25 psi, as is the case when low pressure devices are employed to reduce gas surge?

Answer: Comparative flow tests were made with a typical system having a delivery pressure of 25 psi (Photo Left) and a commercial low pressure system (Photo Right) that required a pressure of only 9 psi to flow the 31 CFH.  This particular low pressure system mounts at the wire feeder and includes a flow calibrated pressure gauge. 

In these tests both devices were set at 31 CFH with a restriction in the feeder/torch system requiring 5 psi to flow that amount of shielding gas.  Restrictions were varied simulating spatter build-up in the torch nozzle, partially blocked gas diffuser, twisted cables and debris accumulation in the wire conduit which is often used as a gas passage in the torch cable.

 The restriction pressure drop varied from 3 psi to 8 psi.  The flow settings in both cases were not changed, as if  flow controls were locked.

See flow results in table below:

As noted, the flow control device with upstream pressure exceeding 25 psi held the same preset flow of 31 CFH as the restrictions were varied from 3 to 8 psi.

With the low pressure system, without any change in flow setting and with the flow calibrated pressure gauge reading a constant 31 CFH - the flow actually varied from a high of 37 CFH to a low of 16 CFH!  This range is much larger than any acceptable Welding Procedure Specification would allow. 

The flow gauge read a constant 31 CFH while the flow varied from 37 to 16 CFH since it is not actually measuring flow, it measures the pressure above an orifice.  In this instance since the pressure is not above 25 psi, the flow though the orifice is not a critical.  Therefore the flow varies with downstream pressure changes caused by added restrictions.

Some low pressure ad-on devices use even lower pressures and can create even larger flow variations.  The one shown in the photo right (yellow arrow) attaches to a standard flowmeter and sets flow by setting pressure.  Where the low pressure device test mentioned above changed 4 CFH with a 1 psi increase in flow resistance; in tests of this device the equivalent of a 1 psi increase in resistance caused the flow to reduce 8 CFH!  Unfortunately some fabricators who have tried and rejected these devices are fearful our Gas Saver System (GSS) may cause similar problems.  Our patented GSS  does not alter pressures and will not cause flow variations!  See Overview of GSS

BE CAREFUL SOME OF THESE DEVICES DO NOT MENTION THAT THEY LOWER PRESSURE!!  ALSO ANY DEVICE THAT MOUNTS DIRECTLY AT THE FEEDER WILL NOT PROVIDE SUFFICIENT EXTRA START GAS.  EMAIL IF YOU HAVE A QUESTION ABOUT A SPECIFIC MODEL.

Question:  Are there production examples where low pressure devices have caused problems?

Answer: We have encountered a number of examples where this problem has been observed and the devices removed and discarded.

Quoting One Specific Case.  A Welding Engineer at a large automotive parts supplier installed low pressure devices that mounted at the feeder and included a pressure calibrated flow gauge (the same device used in the tests mentioned above.)  Here are his quotes about the experience:

“After purchasing and using 32 low pressure gas saving devices (Photo Right) that mounted at the wire feeders we decided to discard all of them!  There were two major problems:

1) Lack of sufficient extra gas at the start made inferior starts and

2) Large flow variations from preset levels were evident when flow was checked at the torch.  In fact as he stated; "Even if the flow was blocked, the flow calibrated pressure gauge supplied with these devices had the same preset reading!"

A Second Case. A Welding Engineer at a major Mid West heavy construction equipment company was having porosity in the same weldment in one plant and not in another.  He performed a very careful, systematic  analysis of the problem.  A fishbone troubleshooting diagram was developed with over 30 items considered in attempt to solving the problem.  It appeared to be a nitrogen porosity problem so he looked at all the possible causes.  They checked for pipeline, solenoid, feeder plumbing leaks, putting on cylinder gas supply to validate the quality of shielding gas and many others items.  He even tried a different type and manufacturer of the solid wire they were using. 

After all this testing and elimination of differences he found one major cause was the plant with the porosity problem was using low pressure surge reducing devices on their welders!  The same device as shown in the report mentioned above. These were removed and the problems went away!!   Variability in flow was probably allowing nitrogen to enter the shielding gas stream.  If shielding gas flow rate is too high or too low this can occur. 

We find fabricators often use too high a flow rate which after about 50 CFH with a typical 5/8 inch ID nozzle are just pulling air with it's 70% nitrogen and moisture into the weld due to turbulence in the shielding stream.

A Third Documented Case: When evaluating the fittings they would need for adding GSS's to 45 MIG welders at a bar joist manufacturer, it was observed that one installation had a low pressure regulator/flow control device mounted at one feeder (same device shown in case 1 and used in case 2).  It had an output pressure gauge calibrated in CFH.  The maintenance manager indicated these had been installed on all welders several years before at a cost much higher than the GSS  and had to be removed due to inconsistent results!  The one remaining was adjusted at a very high flow rate according to the flow gauge reading. The maintenance manager indicated it would not be reinstalled and was surprised it was still there!

Bottom Line: Low pressure at first appears to offer a solution to gas surge, as do restriction orifices mounted at the feeder.  But it creates other larger problems!

You'll note his comment about lack of sufficient extra gas making inferior starts in one of the reported problem cases.  That will occur with any device that controls gas flow at the feeder, low pressure devices, flowmeters or simple orifices.  We have seen a number of problems caused by the use of this flow control location.