The availability of modern ultra fast in-situ gob temperature measurements proved that there is lot going on in between the nine-grid and the orifice. Since it is of outmost importance for the forming process to obtain equal gob temperatures in seems to be obvious to look after different ways of controlling the glass running into the spout and the way it is transported in the spout to the orifice. Plungers and tube control proved to have a big influence on the temperature distribution over the gobs hence they are still not actively controlled by any systems other then gob weight control.
Managing equal gob temperatures in a double gob forming process can be done without major issue. However, the gob temperature stability is not guaranteed by the nine-grid stability.
“Controlling equal gob temperature in triple gob or quad-gob processes becomes exponentially more difficult”.
To overcome some of those problems, thicker middle plungers are normally used to be able to produce shorter middle gob(s) and move the corresponding shear blades closer to the orifice. But in fact this is only a work-around to overcome different gob temperature levels.
Recent tests with ultra fast optical gob temperature measurements showed that there is a potential difference in between the three gobs in a triple gob NNPB process which wasn’t indicated by the nine-grid. However, averaging the three optical measured temperatures showed a clear correlation with the nine-grid thermo element measurements indicating that both nine-grid and optical gob measurements are following the same glass temperature fluctuations. The average optical gob temperatures, following the nine grid fluctuations clearly prove the plausibility of such new measurements.
Nine grid temperature based control strategies are well known and if the forehearth is well designed and running at its designed pull rate, normal PID control is capable of controlling the nine grid sufficiently. The basic rule of thumb has not changed over the years; “control less and stability of furnace overrules preciseness of forehearth temperature control”. More sophisticated, interacting PID controls or even (regious) model based controllers can do without operator interference but will add more cost to the control system. Most probably those extra costs will only be justified in a frequent job change situation.
To conclude, it is very difficult to influence or even control individual gob temperature from out of the forehearth and the nine grid does not give sufficient information since it is placed before the spout, not taking into account all fluid and temperature dynamics of the spout, tube, plungers and orifice.
Having the gob temperatures available these measurements are an important instrument to determine the spout dynamics. Filtered and applied in the correct way to a controller gives us some potential improvements on gob temperature equality.
Control strategy approach
Quality of the nine grid
Dominant on the nine grid seems to be its stability. As long as the nine gird temperatures are stable it shows at least that the forehearth is operating stable and is capable of controlling incoming glass temperature fluctuation coming from other line’s job changes, furnace instabilities, outside temperature fluctuations or building internal draft issues.
Well tuned PID controllers and proper forehearth control strategies, either by operators or automated should be in place but don’t need to try to keep on controlling until all temperature in the nine grid are the same. Again, there is no predefined rule that the nine-grid efficiency needs to be at its optimum. The fact that there are several different calculations to define the nine-grid efficiency show already that there is not such a thing as a “golden rule”.
Expectable equality of temperatures together with best stability should be enough.
Having the nine-grid temperatures available the system should provide all known nine-grid efficiency calculations but of course also a “traffic light” functionality showing a stability figure which might could be achieved by comparing a long history average against a short history average. In fact a comparison of what the forehearth is capable of against how it is performing at the very moment. An underlying trend could show that performance more precise.
Optical measurement plausibility
Having both the nine grid and optical temperature measurement available it will give us, first of all the possibility to perform some plausibility checks on the optical sensors to make sure that the readings have a specific correlation with the nine grid. Correlation or non-correlation could be indicated to the operator by a simple “traffic light” indicator with an underlying trend page to clarify possible courses.
The math behind this plausibility control needs to find out the quality of correlation, regardless the absolute temperature differences, between the average nine grid temperatures, the controlled middle upper nine grid temperature against the average optical gob temperature. The traffic light will indicate a bad, average and good correlation. The trend display needs to show the averaged temperatures on top of each other with temperature differences filtered out.
Ways of control
The principal of gob equality control has to be found in the spout not in the forehearth. Several different sources proved that principal clearly. The temperature drop and gob’s equality temperature in between the nine-grid and the measured gob temperatures must come from the influence of spout equipment like the spout heating, which is normally not controlled, the tube position, speed and rotation direction and plungers. The most feasible CV’s (controlled values) are the spout heating and the tube rotation speed and direction. Closing the possible control loops in between the gob temperature and the spout is one of the improvements Eurotherm is studying right now.