201102149 六、發明說明： 【發明所屬之技術領域】 本發明係關於一種使用於貫通孔内填充非壓縮性纖維 來捕集粒子之慣性過濾器之粒子分級裝置。 【先前技術】 級聯衝擊器（cascade impactor )型粒子分級裝置，係 將衝擊器由上下方向複數段串聯而成之裝置（參照專利文 獻1之圖1，以下稱作第i習知例）。於第丨習知例中，由 於越向下段側則越減小衝擊器中之氣流通過嘴直徑，因此 依序k向氟流速度，而可依序藉由各段衝擊器自慣性質量 較大之粒徑大的粒子中進行捕集分級。衝擊器係使於改變 氣流方向時因慣性力而無法追隨該氣流方向變化之具有慣 性質量之粒子撞擊捕集板而捕集之裝置，於第丨習知例中， 可藉由將衝擊器複數段連接，而按照粒徑之大小依序將粒 子分級。 於具有此種構成之第1習知例中，藉由泵抽吸，降低 裝置内部壓力’藉由裝置内部壓力與裝置外部壓力間之氣 壓差而於裝置内部產生氣流，藉由該氣流使粒子分級。 然而’當粒徑微小之粒子時’由於高精度地製作喷嘴直徑 較為困難，故微小粒子之分級較為困難。 於專利文獻1之圖2中揭示有以下構成：將第】習知 例配置於上段側，而進行粒徑較大之粒子的分級，另一方 面’於下段側配置慣性過濾器，而可謀求微小粒子之分級 201102149 (以下，稱作第2習知例）。第2習知例包含：具有氣體 通過之貫通孔之過濾器支持部、及為堵塞該貫通孔而配置 於該貫通孔内之通氣性多孔質構件即SUS纖維。 [專利文獻1]日本特開2008-70222號公報 【發明内容】 [發明所欲解決之問題] 然而，於第2習知例之構造中，於藉由提高貫通孔内 金屬纖維之填充率來謀求貫通孔内之空隙率之微小化，而 欲進行微小粒子之分級時，貫通孔内氣流流通性會較大降 低而壓力損失增大。其結果，需要作為氣流抽吸泵之大型 設備’而導致攜帶不方便之裝置整體的大型化。另一方面， 考慮到攜帶性而以小流量之小型泵進行氣流抽吸時，於空 隙率微小之貫通孔内因壓力損失而導致氣流速度降低，並 且分級所需要之粒子慣性效果降低，而使目標微小粒子之 分級變得困難。 本發明係可謀求小型輕量化並可實現微小粒子之分級 作為應解決之課題。 [解決問題之技術手段] 本發明之粒子分級裝置係： 自流體流路之上游側至下游側串聯連結配置有至少2 個慣性過濾器； 201102149 使位於上游側的第丨慣性過遽器之該非壓縮性纖維之 纖維徑，大於位於下游側的第2該慣性過據器之該非壓縮 性纖維之纖維徑； 將該第1慣性過濾器作為粗粒子除去過濾器丨 將該第2慣性過濾器作為微小粒子分級過濾器。 根據本發明，於第1慣性過遽器中，於調整纖維徑較 大之非壓、缩性纖維之填充量而將其壓力#失抑制在儘可能 小之狀態下，可有效地進行粒徑較大之粒子的捕集。又， 於第2償性過濾器中’於調整纖維徑較小之非壓縮性纖維 之充填1而將其壓力損失抑制在儘可能小之狀態下，可有 效地進行粒徑較小之粒子的捕集。 藉此，作為第1、第2慣性過濾器中之氣流抽吸構成， 即便採用小型輕量、低抽吸流量之泵’亦可於將壓力損失 抑制在儘可能小之狀態下，自粒徑較大之粒子中分離捕集 粒徑較小之粒子。 本發明中，例如將第丨慣性過濾器之貫通孔内的非壓 縮性纖維之纖維徑dl設為12/am時，雖然壓力損失與非壓 縮I"生’截維之填充量增加大致成正比例地增加，但於流量每 分鐘6公升之小流量時，於將其壓力損失抑制在0.2-0.5 Pa 級之極小之狀態下，可進行將粒徑0.5 v m左右作為分離徑 之粗粒子之捕集（具有大於粒徑0·5 M m之粒徑的粗粒子之 捕集）。又，將第2慣性過濾器之貫通孔内的非壓縮性纖 維之纖維徑dl設為8 μ m時，於流量每分鐘6公升之小流 量時’即便使非壓縮性纖維之填充量增加，壓力損失亦不 201102149 會與其成正比例地增加’於將壓力損力抑制在〇 · 2 _ 〇. $ p a級 之極小之狀態下，可進行將粒徑約190nm&右作為分離徑 之微少粒子之捕集（具有小於粒徑約19〇 nm之粒徑的微少 粒子之捕集）。其結果，即便以小流量之泵進行抽吸亦可 獲得分級所需要的粒子慣性效果’從而可謀求裝置整體之 小型輕量化。 [發明之效果] 根據本發明，可提供一種可謀求系統整體及粒子分級 裝置本身之小型輕量化、並且可於低壓力損失狀態下進行 粗粒子除去及微少粒子之分級的粒子分級裝置。 【實施方式】 以下，參照隨附圖式’對本發明之實施形態之慣性過 濾器及使用其之粒子分級裝置進行說明。再者於實施形 態中假定粒子為漂浮於作為溶劑之—例的氣體中之粒子， 但並不限；t於漂浮於氣體中之粒子，彳包含漂浮於其他溶 劑例如液體中或其他中之粒子。參照圖1及圖2A、圖2B， 實施形態之粒子分級奘署1 6 + 裝置1自軋流上游側至氣流下游側依 序具備：作為預慣性過潘II夕# , . 貝f過'愿盗之第1慣性過濾器3、作為正式 慣性過濾器之第2慣性過、请哭ς w ^ 丨員r玍過應器5、備用過濾器7 '及將慣性 過濾器内部氣流排出至外AR夕A _ 芏外之排氣部9。第1慣性過濾器3 為粗粒子除去用之過遽器。 等微少粒子分級用之過渡器 第2慣性過濾器5為奈米粒子 。備用過濾器7為微少粒子捕 集用之過濾器。再者， 所謂奈米粒子係指奈米級粒子。 6 201102149 第1慣性過渡器3包含圓板狀板3a、圓筒狀板3b及圓 柱狀板3c ’於該等板3a、3b、3c之内部形成過濾器空間3d。 圓板狀板3a作為過濾器板配置於氣流上游側。圓板狀 板3a具有大量未圖示之氣流吸入孔，藉由配置於氣流下游 側之未圖示之氣流抽吸泵的作用，可自該氣流吸入孔將氣 μ及入至裝置内部。圓板狀板3 a不一定為必需，亦可省略。 圓筒狀板3b具有與圓板狀板3a之外徑相同之外徑，而 構成第1慣性過濾器3之側面。圓柱狀板3c於板中央具有 沿軸方向之貫通孔3e。貫通孔3e具備：自氣流上游側向下 游側内徑逐漸連續地縮徑之縮徑貫通孔3d、及於縮徑貫通 孔3el之下端連續形成為内徑固定之定徑貫通孔3e2。 於定徑貫通孔3e2中，即便通過高速氣流體積亦幾乎無 變化之非壓縮性纖維丨丨以緻密纏繞之狀態填充。非壓縮性 纖維11較佳為由SUS (不鏽鋼）纖維等金屬纖維所構成。 又，作為金屬纖維，並不限定於sus纖維，亦可為選自銘 纖維' 銅纖維、其他金屬纖維之丨種以上之金屬纖維。又， 非壓縮性纖維11若為非壓縮性且即便通過高速氣流體積亦 幾乎無變化之纖維，則並不限定於金屬纖維。 第2慣性過濾器5係連續配置於第卜償性過滤器3之 氣流下游側，而連接於第i慣性過濾器3。第2慣性過濾器 匕3八有與第1慣性過濾器3之外徑相同的外徑之圓筒狀 板5a、及圓柱狀板5b，於該等板之内部構成過滤器 空間5c。 圓柱狀板5b於板中央具有沿軸方向之貫通孔…貫通 201102149 孔5d具備：自氣流上游側向下游側内徑逐漸連續地縮徑之 縮徑貫通孔5dl '及於縮徑貫通孔部分5dl之下端連續形成 為内徑固定之定徑貫通孔5d2。 於定徑貫通孔5d2中，即便通過高速氣流體積亦幾乎 無變化之非壓縮性纖維13以緻密纏繞之狀態填充。非壓縮 性纖維1 3較佳為由SUS(不鏽鋼）纖維等金屬纖維所構成。 再者，作為金屬纖維，並不限定於SUS纖維，亦可為選自 銘纖維、銅纖維 '其他金屬纖維之1種以上之金屬纖維。 又，非壓縮性纖維13若為非壓縮性且即便通過高速氣流體 積亦幾乎無變化之纖維，則並不限定於金屬纖維。 備用過濾器7係連續配置於第2慣性過濾器5之氣漭 下游側，而連接於帛2慣性過濾1 5。備帛過滤°器7包含二 具有與第2慣性過濾H 5 <外徑相同的外徑之圓筒狀板 h、及圓板狀板7b。圓板狀出作為過渡器板之作 用。於該等板7a、7b之内部形成過濾器空間九。 排氣部9係自上述裝置内向外部排出氣流者，藉由未 圖示之柚吸泵進行上述排氣。 於以上構成中，於氣流如箭頭所示般自氣流上游側之 第1慣性過滤器3向氣流下游側之排氣部9流動，並通過 各過渡器3、5、7時，粗粒子由第1過據器3除去，而且 微少粒子由第2慣性過滤器5分級，經分級之微少粒子由 備用過濾器7捕集。 於本發明之實施形態中， 貫通孔3e2與第2慣性過濾器 於第1慣性過濾器3之定徑 之疋彳里貫通孔5d2之各定徑 201102149 貫通孔中填充非壓縮性纖維丨丨、丨3 ^該非壓縮性纖維η、 13於本實施形態中為SUS纖維。 若將定徑貫通孔3e2内之非壓縮性纖維丨丨之纖維徑（" m)設為dl ’將定徑貫通孔5d2内之非壓縮性纖維13之纖 維徑（Μπ〇設為d2，則該等有dl>d2之關係。又，將定 徑貫通孔3e2、定徑貫通孔5d2各自之非壓縮性纖維n、13 之填充里（mg )設為mi、m2，將與該非壓縮性纖維填充量 nU'm2對應之壓力損失（kpa :千帕斯卡）分別設為△”、 △ Ρ2。 於以上之構成中，第丨慣性過濾器3之縮徑貫通孔3el 越向氣流下游側方向，直徑變得越小，故氣流緩慢加速後， 以一定速度通過定徑貫通孔3e2，於通過時捕集粗粒子。 由於定徑貫通孔3e2變成非壓縮性纖維丨丨為層狀之過 濾器構造，故可應用在氣體之流速、纖維徑之選擇中能使 用的史托克斯（Stokes)數Stk、及貝克勒（peciet)數Pe。 史托克斯數Stk係表示於非壓縮性纖維構造之過濾器内粒 子對氣體流動之追從性的無因次之值。其式省略。史托克 斯數Stk與流速及粒子密度成正比例，與粒徑之平方成正比 例’與纖維徑成反比例。 根據史托克斯數Stk之式，隨著氣體之流速變大，自粒 徑較大之漂浮粒子起無法有序地追從氣體之運動，而自氣 體之流路偏離與金屬纖維碰撞。參考該史托克斯數Stk且控 制氣體之流速及選擇纖維徑，藉此可選擇捕集目標粒子之 粒徑。由於實施形態中金屬纖維之纖維徑極小，故無須增 201102149 大衝擊器之程度的流速。又，非壓縮性纖維（特別是金屬 纖維）不僅藉由粒子之慣性，而且藉由使用遮擋、$力、 靜電力 '擴散等捕集機構亦可捕集粒子。 貝克勒數Pe係表示粒子因氣流而搬送之效果與粒子因 擴散而搬送之效果的比率之數，與流速、纖維徑成正比例， 與擴散係數成反比例。為減少擴散之影響，而必須增大貝 克勒數Pe。粒徑越小，擴散係數會變得越大，纖維徑可選 擇越小之值，因此可知提高流速會提高粒徑之選擇性而較 佳。基於以上敍述，藉由選擇流速、纖維徑等，而可藉由 非壓縮性纖維捕集或分級目標粒子。 本實施形態中，於第1慣性過濾器3中，藉由調整定 徑貫通孔3e2中之空隙率（具體而言，調整定徑貫通孔3e2 中非壓縮性纖維11之填充量）、以及設定非壓縮性纖維u 之纖維徑dl，而可抑制（減小）壓力損失而不會使定徑貫 通孔3e2内之氣流流通性較大降低。其結果，即便以小型氣 流抽吸泵進行小流量抽吸，亦可獲得粗粒子除去所必需的 粒子慣性效果。 同樣’於第2慣性過濾器5中，藉由調整定徑貫通孔 5d2中之空隙率（具體而言，調整定徑貫通孔5d2中非壓縮 性纖維13之填充量）、以及設定非壓縮性纖維13之纖維 & d2 ’而可抑制（減小）壓力損失而不會使定徑貫通孔3d2 内之氣流流通性較大降低。其結果，即便以小型氣流抽吸 泉進行小流量抽吸’亦可獲得微少粒子分級所必需的粒子 慣性效果。 10 201102149 例如減小非壓縮性纖維11、13之填充率而增大定徑貫 通孔3e2、5d2内之空隙率，而且選擇較小的非壓縮性纖維 11、13之直徑dl、d2。藉此’不會使定徑貫通孔3e2、5d2 内之氣流流通性較大降低，可獲得微少粒子分級所必需的 粒子慣性效果。其結果，即便以小型氣流抽吸栗進行小流 量抽吸，亦可抑制壓力損失並進行微少粒子分級。 對上述實施形態中之具體的數值例進行說明。於定徑 貫通孔3e2與定徑貫通孔5d2中，將其孔徑D1、D2設為3 mm、6 mm，將其孔長Li、L2設為4.5 mm、5 mm。將非壓 縮性纖維11、13之纖維徑dl、d2分別設為12//m、8#me 又，藉由利用氣流抽吸泵進行抽吸，藉此產生之氣流之流 量Q1、Q 2均為小流量，即每分鐘6公升。 將根據以上條件之金屬纖維填充量m 1、m2與壓力損失 △P1、ΛΡ2之關係示於圖3A、圖3B。於圖3A中，橫軸為 疋徑貫通孔3e2内之非壓縮性纖維之填充量ml ( mg )，縱 軸為壓力損失ΛΡ1 (kPa)。於圖3B中，橫軸為定徑貫通 孔5d2内之非壓縮性纖維之填充量m2(mg)，縱軸為壓力 損失 ΛΡ2 ( kPa)。 如圖3A、圖3B所示，於非壓縮性纖維填充量mi為 1〇_2〇 mg之調整範圍内，定徑貫通孔3e2之壓力損失 為〇.3-0·4 kPa，於非壓縮性纖維填充量m2為3-4 mg之調 整範圍内，定徑貫通孔5d2之壓力損失△”為i 5·2α&。 。如由圖3A、圖3B之資料可明確，即便一面實施藉由 可攜式菜亦可謀求之小流量抽吸，一面將帛"貫性過渡器 11 201102149 3、第2慣性過濾器5内夕此麻，, 巧之非壓縮性纖維11、13各自之埴 充量（空隙率）在其填充I >田妓# m 兄量调整範圍内進行調整，亦可 低壓力損失狀態下，進行m ]，睹丨止,各μ 、 仃第1慣性過濾器3中之粗粒子除 去、及第2慣性過濾器5巾> ^ ^ π、 示 中之微少粒子分級。因此， 亦可藉由小型輕量之粒子公铋腚 _ 刀級裝置，而尚精度地測定工 者之呼吸區域中之微小粒子曝露量。 圖4A、4B分別表元笛1啤lL、 ^ 墉51 S由Π 過濾器3、及第2慣性過 〜中之粒徑（心)，集效率（％)之關係、。該等圖中， 橫轴為粒徑（P)，縱轴為粒子捕集效率（％)。 第1慣性過濾器3、第2憎枓讲油C A i 、 閬性過濾益1各自之定徑貫通孔 3e2、5d2 之孔徑 Dl、D2 盔 ^BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a particle classifying device for an inertial filter for trapping particles in a through-hole filled with non-compressible fibers. [Prior Art] A cascade impactor type particle classifying device is a device in which a plurality of stages of the impactor are connected in series from the vertical direction (see Fig. 1 of Patent Document 1, hereinafter referred to as the first example). In the second example, the lower the downstream side, the more the airflow in the impactor is reduced through the nozzle diameter, so that the velocity of the fluorine flow is sequentially increased, and the self-inertial mass of each of the impactors can be sequentially large. The particles having a large particle size are collected and classified. The impact device is a device that captures a particle having an inertial mass that is incapable of following the change in the direction of the airflow when it changes the direction of the airflow, and the device is captured by the capture plate. In the second example, the impactor can be multiplied by the impactor. The segments are connected, and the particles are sequentially classified according to the size of the particle size. In the first conventional example having such a configuration, the internal pressure of the apparatus is lowered by pumping, and a gas flow is generated inside the apparatus by the difference in pressure between the internal pressure of the apparatus and the external pressure of the apparatus, and the airflow is used to cause the particles. Grading. However, when the particles having a small particle size are difficult to fabricate the nozzle diameter with high precision, it is difficult to classify the fine particles. In Fig. 2 of Patent Document 1, there is disclosed a configuration in which the first conventional example is disposed on the upper side to classify particles having a large particle size, and the inertial filter is disposed on the lower side. The classification of the fine particles is 201102149 (hereinafter referred to as the second conventional example). The second conventional example includes a filter support portion having a through hole through which a gas passes, and a SUS fiber which is an air permeable porous member which is disposed in the through hole to block the through hole. [Patent Document 1] JP-A-2008-70222 SUMMARY OF INVENTION [Problems to be Solved by the Invention] However, in the structure of the second conventional example, the filling rate of the metal fibers in the through holes is increased. When the porosity of the through-hole is miniaturized, and the classification of the fine particles is desired, the gas flowability in the through-hole is largely lowered, and the pressure loss is increased. As a result, it is necessary to use a large-scale apparatus as an air suction pump, which results in an increase in size of the entire device that is inconvenient to carry. On the other hand, when the airflow is sucked by a small pump having a small flow rate in consideration of the portability, the airflow speed is lowered due to the pressure loss in the through hole having a small void ratio, and the particle inertia effect required for the classification is lowered, and the target is lowered. The classification of tiny particles becomes difficult. According to the present invention, it is possible to achieve compactness and weight reduction and to realize classification of fine particles as a problem to be solved. [Means for Solving the Problem] The particle classifying device of the present invention is configured such that at least two inertial filters are connected in series from the upstream side to the downstream side of the fluid flow path; 201102149 makes the non-inertial inertia device located on the upstream side The fiber diameter of the compressible fiber is larger than the fiber diameter of the non-compressible fiber of the second inertial filter located on the downstream side; the first inertial filter is used as a coarse particle removal filter, and the second inertial filter is used as the second inertial filter Micro particle grading filter. According to the present invention, in the first inertia filter, the amount of the non-pressure-reducing fibers having a large fiber diameter is adjusted, and the pressure # is suppressed as small as possible, whereby the particle diameter can be effectively performed. Capture of larger particles. Further, in the second compensating filter, the pressure loss of the non-compressible fiber having a small fiber diameter is adjusted to suppress the pressure loss as small as possible, and the particles having a small particle diameter can be efficiently produced. Capture. Therefore, as the airflow suction configuration in the first and second inertia filters, even if a small-sized, lightweight, low-suction flow pump can be used, the pressure loss can be suppressed as small as possible. Separate and trap particles with smaller particle sizes in larger particles. In the present invention, for example, when the fiber diameter dl of the non-compressible fiber in the through hole of the second inertial filter is 12/am, the pressure loss is substantially proportional to the increase in the filling amount of the uncompressed I" When the flow rate is increased, the flow rate is reduced to a minimum of 0.2-0.5 Pa at a small flow rate of 6 liters per minute, and the capture of coarse particles having a particle diameter of about 0.5 vm as a separation diameter can be performed. (Capture of coarse particles having a particle size larger than the particle size of 0·5 M m). In addition, when the fiber diameter dl of the incompressible fiber in the through hole of the second inertia filter is 8 μm, the amount of the non-compressible fiber is increased even when the flow rate is small at a flow rate of 6 liters per minute. The pressure loss is not increased by 201102149. In the state where the pressure loss is suppressed to a minimum of 〇· 2 _ 〇. $ pa, a particle with a particle size of about 190 nm & Capture (capture of tiny particles having a particle size smaller than about 19 〇 nm in particle size). As a result, even if the pump is pumped with a small flow rate, the particle inertia effect required for classification can be obtained, and the overall size and weight of the device can be reduced. [Effects of the Invention] According to the present invention, it is possible to provide a particle classifying device which can reduce the size and weight of the entire system and the particle classifying device itself, and can perform coarse particle removal and fine particle classification in a low pressure loss state. [Embodiment] Hereinafter, an inertial filter according to an embodiment of the present invention and a particle classifying device using the same will be described with reference to the accompanying drawings. Further, in the embodiment, it is assumed that the particles are particles floating in a gas as a solvent, but are not limited; t are particles floating in a gas, and the particles include particles floating in other solvents such as liquid or others. . Referring to Fig. 1 and Fig. 2A and Fig. 2B, the particle grading system 1 6 + device 1 of the embodiment is provided in order from the upstream side of the rolling flow to the downstream side of the air flow: as a pre-inertia over Pan II eve # , . The first inertia filter 3, the second inertia of the official inertia filter, please cried w ^ 玍r玍玍5, the backup filter 7' and the internal airflow of the inertia filter to the outside AR夕 A _ The exhaust part 9 outside the 。. The first inertial filter 3 is a filter for removing coarse particles. The transition device for fine particle grading The second inertial filter 5 is a nanoparticle. The backup filter 7 is a filter for minute particle collection. Furthermore, the so-called nanoparticle refers to a nanoparticle. 6 201102149 The first inertia transition device 3 includes a disk-shaped plate 3a, a cylindrical plate 3b, and a cylindrical plate 3c'. The filter space 3d is formed inside the plates 3a, 3b, and 3c. The disk-shaped plate 3a is disposed as a filter plate on the upstream side of the airflow. The disk-shaped plate 3a has a large number of airflow suction holes (not shown), and the gas suction pump (not shown) disposed on the downstream side of the airflow allows the gas to be introduced into the device from the airflow suction hole. The disc-shaped plate 3 a is not necessarily required and may be omitted. The cylindrical plate 3b has the same outer diameter as the outer diameter of the disk-shaped plate 3a, and constitutes the side surface of the first inertial filter 3. The cylindrical plate 3c has a through hole 3e in the axial direction at the center of the plate. The through hole 3e is provided with a reduced diameter through hole 3d which is gradually reduced in diameter from the upstream side of the airflow toward the downstream side, and a fixed diameter through hole 3e2 which is continuously formed at the lower end of the reduced diameter through hole 3el. In the sizing through hole 3e2, the incompressible fiber bundle which is hardly changed by the high-speed airflow volume is filled in a state of being densely wound. The non-compressible fiber 11 is preferably made of a metal fiber such as SUS (stainless steel) fiber. Further, the metal fiber is not limited to the sus fiber, and may be a metal fiber selected from the group consisting of the "fiber" and the other metal fibers. Further, the non-compressible fiber 11 is not limited to the metal fiber unless it is non-compressible and has almost no change in the volume of the high-speed gas stream. The second inertial filter 5 is continuously disposed on the downstream side of the airflow of the vehicular filter 3, and is connected to the i-th inertia filter 3. The second inertial filter 匕3 has a cylindrical plate 5a having the same outer diameter as the outer diameter of the first inertial filter 3, and a cylindrical plate 5b, and the filter space 5c is formed inside the plates. The cylindrical plate 5b has a through hole in the axial direction at the center of the plate. The through hole 201102149 includes a reduced diameter through hole 5d1 which is gradually reduced in diameter from the upstream side to the downstream side of the airflow side, and a reduced diameter through hole portion 5d1. The lower end is continuously formed into a fixed diameter through hole 5d2 whose inner diameter is fixed. In the sizing through hole 5d2, the incompressible fiber 13 which is hardly changed by the high-speed airflow volume is filled in a state of being densely wound. The non-compressible fiber 13 is preferably made of a metal fiber such as SUS (stainless steel) fiber. In addition, the metal fiber is not limited to the SUS fiber, and may be one or more metal fibers selected from the group consisting of the name fiber and the copper fiber. Further, the non-compressible fiber 13 is not limited to the metal fiber unless it is non-compressible and has almost no change even if it is fluidized by high-speed gas. The backup filter 7 is continuously disposed on the downstream side of the second inertia filter 5, and is connected to the 帛2 inertial filter 15 . The preparation filter 7 includes two cylindrical plates h having the same outer diameter as the outer diameter of the second inertial filter H 5 < and a disk-shaped plate 7b. The disc shape acts as a transition plate. A filter space nine is formed inside the plates 7a, 7b. The exhaust unit 9 discharges the airflow from the inside of the apparatus to the outside, and the exhaust is performed by a pomelo suction pump (not shown). In the above configuration, when the airflow flows from the first inertial filter 3 on the upstream side of the airflow to the exhaust portion 9 on the downstream side of the airflow as indicated by the arrow, and passes through the transitioners 3, 5, and 7, the coarse particles are 1 The passer 3 is removed, and the minute particles are classified by the second inertial filter 5, and the finely divided particles are collected by the backup filter 7. In the embodiment of the present invention, the through hole 3e2 and the second inertia filter are filled with the non-compressive fiber bundle in the through hole of each of the through holes 5d2 of the fixed diameter of the first inertia filter 3;丨3 ^ The non-compressible fibers η and 13 are SUS fibers in the present embodiment. When the fiber diameter (" m) of the incompressible fiber bundle in the fixed diameter through hole 3e2 is dl', the fiber diameter of the incompressible fiber 13 in the fixed diameter through hole 5d2 is set to d2, Then, there is a relationship of dl > d2. Further, the filling (mg) of the incompressible fibers n and 13 of each of the fixed diameter through hole 3e2 and the fixed through hole 5d2 is set to mi and m2, and the incompressibility is obtained. The pressure loss (kpa: kPa) corresponding to the fiber filling amount nU'm2 is Δ" and Δ Ρ2, respectively. In the above configuration, the diameter reducing through hole 3el of the third inertial filter 3 is directed toward the downstream side of the airflow. The smaller the diameter is, the slower the airflow is, and the smaller the through-holes 3e2 are passed through at a constant speed, and the coarse particles are collected during the passage. The fixed-diameter through-holes 3e2 become non-compressive fibers and the layered filter structure is formed. Therefore, the Stokes number Stk and the Peciet number Pe which can be used in the selection of the gas flow rate and the fiber diameter can be applied. The Stokes number Stk is expressed in the non-compressive fiber structure. The dimensionless value of the chasing of the gas flow in the filter. The Stokes number Stk is proportional to the flow rate and the particle density, and is proportional to the square of the particle size' inversely proportional to the fiber diameter. According to the Stokes number Stk, as the gas flow rate becomes larger, The floating particles with larger particle diameters cannot follow the movement of the gas in an orderly manner, and the flow path from the gas deviates from the collision with the metal fibers. Referring to the Stokes number Stk and controlling the flow rate of the gas and selecting the fiber diameter, It is optional to capture the particle size of the target particles. Since the fiber diameter of the metal fiber is extremely small in the embodiment, it is not necessary to increase the flow rate of the large impactor of 201102149. Moreover, the non-compressive fiber (especially the metal fiber) is not only by the particle Inertia, and by using a trapping mechanism such as occlusion, force, and electrostatic force 'diffusion, the particles can also be trapped. The Becker number Pe is the ratio of the effect of the particles being transported by the airflow and the effect of the particles being transported by diffusion. , proportional to the flow rate and fiber diameter, and inversely proportional to the diffusion coefficient. To reduce the influence of diffusion, the Becker number Pe must be increased. The smaller the particle size, the more the diffusion coefficient becomes. The fiber diameter can be selected to be smaller, so that it is better to increase the flow rate by increasing the flow rate. Based on the above description, by selecting the flow rate, the fiber diameter, etc., it is possible to capture or classify by non-compressible fibers. In the first inertia filter 3, the porosity of the fixed diameter through hole 3e2 is adjusted (specifically, the filling amount of the non-compressible fiber 11 in the fixed diameter through hole 3e2 is adjusted) And setting the fiber diameter dl of the non-compressible fiber u, and suppressing (reducing) the pressure loss without greatly reducing the gas flowability in the constant-diameter through hole 3e2. As a result, even a small-sized air suction pump A small flow rate of suction can also achieve the particle inertia effect necessary for coarse particle removal. Similarly, in the second inertial filter 5, the void ratio in the fixed diameter through hole 5d2 (specifically, the filling amount of the non-compressible fiber 13 in the fixed diameter through hole 5d2) is adjusted, and the incompressibility is set. The fiber & d2' of the fiber 13 can suppress (reduce) the pressure loss without greatly reducing the airflow property in the sizing through hole 3d2. As a result, the particle inertia effect necessary for the fine particle classification can be obtained even if the small-flow air suction spring is used for small-flow suction. 10 201102149 For example, the filling ratio of the non-compressible fibers 11 and 13 is reduced to increase the void ratio in the sizing holes 3e2 and 5d2, and the diameters d1 and d2 of the smaller non-compressible fibers 11 and 13 are selected. Thereby, the air flow property in the constant-diameter through holes 3e2 and 5d2 is not greatly reduced, and the particle inertia effect necessary for the fine particle classification can be obtained. As a result, even if the small flow is sucked by the small airflow for small-flow suction, the pressure loss can be suppressed and the particle fractionation can be performed. Specific numerical examples in the above embodiments will be described. In the fixed diameter through hole 3e2 and the fixed diameter through hole 5d2, the apertures D1 and D2 are set to 3 mm and 6 mm, and the hole lengths Li and L2 are set to 4.5 mm and 5 mm. The fiber diameters d1 and d2 of the non-compressible fibers 11 and 13 are set to 12/m and 8#me, respectively, and suction is performed by an air suction pump, whereby the flow rates Q1 and Q2 of the airflow are generated. For small flows, ie 6 liters per minute. The relationship between the metal fiber filling amounts m1 and m2 and the pressure loss ΔP1 and ΛΡ2 according to the above conditions is shown in Figs. 3A and 3B. In Fig. 3A, the horizontal axis represents the filling amount ml (mg) of the incompressible fibers in the diameter through hole 3e2, and the vertical axis represents the pressure loss ΛΡ1 (kPa). In Fig. 3B, the horizontal axis represents the filling amount m2 (mg) of the incompressible fibers in the fixed diameter through hole 5d2, and the vertical axis represents the pressure loss ΛΡ2 (kPa). As shown in FIG. 3A and FIG. 3B, in the adjustment range in which the non-compressible fiber filling amount mi is 1 〇 2 〇 mg, the pressure loss of the sizing through hole 3e2 is 〇.3-0·4 kPa, which is uncompressed. The filling amount m2 is within the adjustment range of 3-4 mg, and the pressure loss Δ" of the sizing through hole 5d2 is i 5 · 2α &. As can be seen from the data of Figs. 3A and 3B, even one side is implemented by The portable dish can also be pumped with a small flow rate, and the 帛"transitional transition device 11 201102149 3, the second inertial filter 5 is numb in the middle, and the incompressible fibers 11 and 13 are the same. The charge (void ratio) is adjusted within the range of the filling I > Tian 妓 # m brother, and can also be performed under the condition of low pressure loss, m], ,, each μ, 仃 first inertia filter 3 The coarse particle removal in the middle, and the second inertial filter 5 towel > ^ ^ π, the microparticles in the display are classified. Therefore, it is also possible to accurately and compactly use the small and lightweight particle public 铋腚 knife-level device. The amount of microparticles exposed in the breathing zone of the worker is determined. Figure 4A, 4B respectively, the table flute 1 beer lL, ^ 墉 51 S by Π filter 3, and 2 The relationship between the particle diameter (heart) and the collection efficiency (%) in the inertia to medium range. In the figures, the horizontal axis represents the particle diameter (P), and the vertical axis represents the particle collection efficiency (%). 3, the second 憎枓 油 oil CA i, 阆 过滤 filter benefits 1 each of the diameter of the through hole 3e2, 5d2 aperture Dl, D2 helmet ^
2為2 3 4 5 6 7 8職、3 mm，孔長u、L 3 mm、4·5 mm，非壓输铋她泌ι， 縮陡纖維"、9之纖維徑心、心為 m ^m°又’藉由利用氣流抽吸泵進行抽吸， 生之氣流之流4Q1、Q2均為小流量（每分鐘6公升）產 如圖4A所示，於第1憎柯 、乐1償性過濾1§ 3中可將〇.5 β 作為粒子分離徑，又，如圖4Β所 12 1 中可H# Θ彳〇η 於第2慣性過濾器 2 甲了將約190 nm作為粒子分離徑。 3 又’亦可以圖5所示之方式構成第卜償性過 4 第2慣性過濾器5。對圖$與 及 5 號來表示。圖5中B賦予相同符 與 形成為使第1慣性過濾器貫通孔3e2 6 、第2慣性過據器書福^丨q4立 貫 連接之形態。該圖5所示之 7 構造亦具有與圖1相同之作用。 之 8 如以上所說明，於本實施形態 流下游你丨夕I· T ^ t於氣"丨L上游側與氣 9 斿側之上下2段，串聯連結配置第 罝弟1、第2慣性過濾器 201102149 3、5,使第Η貫性過遽器3之貫通孔3e2内的非壓縮性纖維 11之纖維徑，大於第2慣性過渡器5之貫通孔5d2内的非 壓縮性纖維13之纖維徑。藉此，於第丨慣性過濾器3中， 由於非壓縮性纖維Π之纖維徑較大，因此即便增多非壓縮 性纖維11之填充量，亦可於將其壓力損失抑制在儘可能小 之狀態下，有效地進行粗粒子之捕集。進而，於第2慣性 過濾器5中，由於非壓縮性纖維13之纖維徑較小，因此即 便減少非壓縮性纖維13之填充量，亦可於將其壓力損失抑 制在儘可能小之狀態下有效地進行微少粒子之分級。藉 此，即便使用小型輕量、低抽吸流量之泵，亦可於將壓力 損失抑制在儘可能小之狀態下，謀求自粗粒子至微少粒子 之分離捕集。 【圖式簡單說明】 圖1係表示自側面觀察本發明之實施形態之粒子分級 裝置的概念構成之圖。 圖2A係表示圖丨裝置内之第丨慣性過濾器之放大圖。 圖2B係表示圖丨裝置内之第2慣性過濾器之放大圖。 圖3A係表示第丨慣性過濾器與第2慣性過濾器各自之 金屬纖維填充量-壓力損失之第1圖。 圖3B係表示第丨慣性過濾器與第2慣性過濾器各自之 金屬纖維填充量-壓力損失之第2圖。 圖4A係表示第丨慣性過濾器及第2慣性過濾器各自之 粒仏（# m)-捕集效率（％ )之關係之第1圖，該圖中橫軸 13 201102149 為粒徑（// m) ’縱軸為粒子捕集效率（％)。 圖4B係表示第丨慣性過濾器及第2慣性過濾器各自之 粒性·（ y m )-捕集效率（〇/ )之關係篦 u , / 關你之第2圖，該圖中橫軸 為粒徑（// m)，縱軸為粒子捕集效率（％)。 、 圖5係表示實施形態之粒子分級裝置的變形 j之圖。 【主要元件符號說明】 粒子分級裝置 第1慣性過濾器 3a、7b 圓板狀板 3b 、 5a 、 7a 圓筒狀板 3c ' 5b 圓柱狀板 3d 、 5c 、 7c 過濾器空間 3e、5d 貫通孔 3el ' 5dl 縮徑貫通孔 3e2 、 5d2 定徑貫通孔 5 第2慣性過濾器 7 備用過濾器 9 排氣部 11、13 非壓縮性纖維 dl > d2 纖維徑 D1、D2 孔徑 LI、L2 孔長 142 is 2 3 4 5 6 7 8 positions, 3 mm, hole length u, L 3 mm, 4·5 mm, non-pressure transmission, her secretion, shrinking fiber ", 9 fiber diameter, heart is m ^m° and 'by using the air suction pump for suction, the flow of the raw gas stream 4Q1, Q2 are small flow (6 liters per minute) production as shown in Figure 4A, in the first 憎 Ke, Le 1 In the case of sexual filtering 1 § 3, 〇.5 β can be used as the particle separation diameter, and, as shown in Fig. 4Β, 12 1 can be H# Θ彳〇η in the second inertial filter 2, and about 190 nm is used as the particle separation diameter. . 3 Further, it is also possible to constitute the second inertial filter 5 in the form shown in Fig. 5. It is represented by the figures $ and and 5. In Fig. 5, B is given the same sign and is formed such that the first inertial filter through hole 3e2 6 and the second inertial filter are connected to each other. The 7 configuration shown in Fig. 5 also has the same effect as Fig. 1. 8 As described above, in the downstream of this embodiment, you are in the middle of the gas and the upper side of the gas and the upper side of the gas 9 斿L, and the second and second inertia are arranged in series. The filter 201102149 3, 5 has a fiber diameter of the non-compressible fiber 11 in the through hole 3e2 of the second pass filter 3 larger than that of the non-compressible fiber 13 in the through hole 5d2 of the second inertia transition device 5. Fiber diameter. Therefore, in the second inertia filter 3, since the fiber diameter of the non-compressible fiber bundle is large, even if the filling amount of the non-compressible fiber 11 is increased, the pressure loss can be suppressed as small as possible. Underneath, the capture of coarse particles is effectively performed. Further, in the second inertial filter 5, since the fiber diameter of the non-compressible fiber 13 is small, even if the filling amount of the non-compressible fiber 13 is reduced, the pressure loss can be suppressed as small as possible. Effectively classify tiny particles. As a result, even if a pump with a small size and a low suction flow rate is used, the pressure loss can be suppressed as small as possible, and separation and collection from coarse particles to minute particles can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a view showing a conceptual configuration of a particle classifying device according to an embodiment of the present invention as seen from the side. Fig. 2A is an enlarged view showing a third inertial filter in the drawing device. Fig. 2B is an enlarged view showing the second inertial filter in the drawing device. Fig. 3A is a first view showing a metal fiber filling amount-pressure loss of each of the second inertial filter and the second inertial filter. Fig. 3B is a second view showing the amount of metal fiber filling-pressure loss of each of the second inertial filter and the second inertial filter. 4A is a first view showing the relationship between the particle size (#m) and the collection efficiency (%) of each of the second inertial filter and the second inertial filter, in which the horizontal axis 13 201102149 is the particle diameter (//). m) 'The vertical axis is the particle capture efficiency (%). 4B is a graph showing the relationship between the graininess (ym)-trapping efficiency (〇/) of the second inertial filter and the second inertial filter, and the second graph in which the horizontal axis is The particle size (//m) and the vertical axis are the particle collection efficiency (%). Fig. 5 is a view showing a modification j of the particle classifying device of the embodiment. [Explanation of main component symbols] Particle classifying device 1st inertia filter 3a, 7b Disc-shaped plate 3b, 5a, 7a Cylindrical plate 3c' 5b Cylindrical plate 3d, 5c, 7c Filter space 3e, 5d Through hole 3el ' 5dl reduced diameter through hole 3e2, 5d2 fixed through hole 5 second inertial filter 7 spare filter 9 exhaust part 11, 13 non-compressible fiber dl > d2 fiber diameter D1, D2 aperture LI, L2 hole length 14