The intak e air ma ss flow rat e is cont inuously mea- sur ed by a hot wire an emometer placed in th e engine intak e man ifold。 Two ident ical press ur e tran sd ucers (JUMO type 4 AÄ P-30; measur ement ran ge -1 to +3 bar) ar e placed at th e inlet an d outlet of th e part iculat e filter to monitor th e press ur e drop。 Temperatur es ar e mea- sur ed simu ltan eously at thr ee locat ions: cata lyst inlet, filter inlet, an d filter out le t using thr ee th ermocouples Ni-Cr-Ni with 0。5 mm  th ickn ess 。

The car bon monoxide concentrat ions ar e measur ed using a nondispe rsive infrar ed (NDIR) ana lyze r (Horiba VIA-510)。 For car bon dioxide, an NDIR ana lyze r (Hart - mann & Braun Utas 3G) is used。 The NO an d NOx concentrat ions ar e measur ed by a chemilum inescence (Signa l series 4000 ) ana lyze r。 The NO2  concentrat ion is calculat ed  by  th e  difference betwee n  NOx  an d  NO

concentrat ions。 Gas ana lysis is perform ed by sam pling before an d after th e part iculat e filter alternat ively using a  thr ee-way valve。

Un ive rsal exhau st gas oxyge n (UE GO) sensors ar e placed ups tr eam of th e cata lyst an d downstr eam of th e filter。 These sensors provide th e air-to-fuel (A/F) ma ss rat io, which is used togeth er with th e intak e air flow to calculat e th e exhau st ma ss flow rat e。 The A/F rat io can be used togeth er with th e exhau st gas ana lysis to calculat e th e O2 cont ent in th e exhau st gas。 The signa ls of th e measuring instrum ent s (th erm ocouples, press ur e tran sd ucers, sensors, an d exhau st ana lyze rs) were recorded in a PC thr ough an A/D convert er。

The fuel used for th ese expe riment s was th e En vi- ronm enta l Class 1 low sulfur cont ent fuel (less than 10 pp m  wt S)。

The tes ts were carr ied out at selected steady-stat e engine operat ing point s, defined by constant engine revolutions an d torque。 The engine-out soot ma ss emis- sions rat es at all of th ese operat ing point s were known from previous measur ement s。 Each steady-stat e operat - ing point was held constant for 9-12 min, before switching to th e next point 。 This time period was sufficie nt to ensur e sta bilizat ion of th e engine operat ing conditions an d th e filter temperatur e。

Exp eri men tal Resu lt s

A typical exam ple of th e ma in recordings dur ing th e expe riment s is prese nt ed in Figur e 2。 Periodically, th e engine is operat ed at a fixed “reference” point (4000 rpm/40 Nm) easily recognized in th e figur e by th e high exhau st flow rat e (>0。08 kg/s)。 The backpr ess ur e mea- sur ed at th ese conditions  can  be  used  for a first assess ment of th e soot loading in th e filter。 Moreover, becau se of th e high soot emissions rat e at th ese condi- tions, th is operat ing point is also used to load th e filter with soot when desired。 The figur e conta ins th e mea- sur ed filter temperatur e, th e exhau st flow rat e, th e NOx an d NO2 concentrat ions at filter inlet, an d th e back- press ur e。 The bottom gra ph is a computat iona l assess - ment of th e instantan eous soot ma ss in th e filter。 This is perform ed, in principle, by using a model-based proced ur e involving an “inverse” solution of th e press ur e drop problem, solving for th e soot ma ss as a function of th e exhau st flow rat e, temperatur e, an d backpr ess ur e。 The press ur e drop model used for th is calculat ion is prese nt ed below。 More deta ils on th is calculat ion pro- ced ur e ar e give n elsew here。19

After initial operat ion at th e reference point , th e engine operat es at thr ee successive operat ing point s (700-2500 s) corr esponding to conditions of relat ively low flow rat e, moderat e-to-high NOx emissions, an d temperatur es ran ging from 290 to 340 °C。  The NO2 cont ent at th e filter inlet is aroun d 250 pp m。 The backpr ess ur e rema ins almost constant at each one of th e thr ee operat ing point s, indicat ing “equilib rium” conditions。 This is also indicat ed by th e es timat ed evolut ion of soot  ma ss, as shown in  th e  bottom gra ph。