Pyrolysis Profile From Run With Walnut Shells - 6/30/09
Our in house testing GEK now has an installed "wand" of thermocouples in the pyrolysis zone, along with thermocouples at "combustion" (1 inch in front of an air nozzle), top of the reduction cone and bottom of the reduction cone (1 inch into the bottom of the cone).
This graph shows the vertical distribution of temperatures (y-axis) with time (x-axis). Start of the graph shows ignition of the reactor.
The color scheme leaves something to be desired, but very small spots of light yellow are above 1000°C, stepping down in 100°C increments.
The wand is comprised of a central 1/8" probe for combustion, with 1/16" TC probes at 1, 2, 3, 4, 6, and 8 inches. A thin stainless tube was also brought down to combustion to monitor the vacuum at that point. The 1/16" probes were spaced apart with alumina beads to limit thermal conduction. The combustion probe was protected with a piece of alumina tubing.
The installed wand.
The R function used to produce the graph:
ContourMap <- function(plot_start=0,plot_end=2000) {
cont<- c(25,100,200,300,400,500,600,700,800,900,1000,1100,1200) cont_col<-c("#FFFFFF","#AA887F","#BB7866","#CC684C","#DD5833","#EE4819","#ff3800","#ff6500","#ff7e00","#ff932c","#ffa54f","#ffb46b","#ffc184") par(xaxt="n") ## don't plot x-axis filled.contour(x=seq(plot_start,plot_end),y=c(-8.5,-2.5,0,1,2,3,4,6,8),z=matrix(data=c(T_bred[plot_start:plot_end],T_tred[plot_start:plot_end],T_comb[plot_start:plot_end],T_1in[plot_start:plot_end],T_2in[plot_start:plot_end],T_3in[plot_start:plot_end],T_4in[plot_start:plot_end],T_6in[plot_start:plot_end],T_8in[plot_start:plot_end]),ncol=9,nrow= (plot_end-plot_start+1)),levels=cont,col=cont_col,xlab="Time [min]",ylab="Position [in]",key.title="T [°C]")
abline(h=0)
abline(h=-2.5)
abline(h=-8.5)
text(x=plot_start,y=0.25,labels="Combustion",pos=4)
text(x=plot_start,y=-2.25,labels="Top Reduction",pos=4)
text(x=plot_start,y=-8.25,labels="Bottom Reduction",pos=4)
par(xaxt="s") ## plot x-axis
axis(1,at=seq(from=plot_start,to=plot_end,by=60*10),labels=seq(from=(plot_start/60),to=(plot_end/60),by=10)) ## plot x-axis in minutes }
Comments (2)
scott Christopher Romack said
at 10:43 pm on Jul 22, 2009
Use these colors
#0000FF,#AA887F,#BB7866,#CC684C,#DD5833,#EE4819,#ff3800,#ff6500,#ff7e00,#ff932c,#ffa54f,#ffb46b,#ffc1;
I can't get the chart to update
bk said
at 11:44 pm on Jul 22, 2009
The code above must be run in R (r-project.org) and requires the dataset from the run, but I figured it may be instructive for others. It looks like your suggesting the blue should be white. Graphically, the tricky part is that one wants a gradient to get a sense of the trend, but then determining which color is which is tricky.
I'll be releasing more of the code to generate these types of plots and other diagnostic graphs, hopefully nearly automatically in R, for other users of the GCU. If anyone is interested in looking in more detail at the code or data, let me know.
You don't have permission to comment on this page.